Trademark Coexistence and Consent Agreements: How Rivals Share a Name and Persuade the USPTO

Summary. This article explains how two businesses that use the same or similar marks can agree to share a name, and why the USPTO and the courts usually respect those agreements. It traces the doctrine from In re E.I. du Pont de Nemours through In re N.A.D., Bongrain, In re Mastic, and In re Four Seasons Hotels to recent Trademark Trial and Appeal Board decisions that accepted or rejected consents, and it explains the difference between a "naked" consent and a "clothed" agreement. It then covers the limits: identifications of goods that do not match the private deal, geographic carve-outs that require a concurrent use proceeding, antitrust rules against dividing markets, naked licensing risk, and the effect on third parties. Separate sections address enforcement between the parties, assignment and successors, bankruptcy after Mission Product Holdings, global coexistence, and tax. An extended worked example follows two companies from a Section 2(d) refusal to a signed agreement and a later acquisition. Common mistakes, an FAQ, and practical takeaways close the article.

Marisol Vega has been general counsel of Tidewell Brewing Company, a craft brewer in Portland, Maine, for six weeks when the office action arrives. Tidewell has applied to register TIDEWELL for canned cocktails, its first product outside beer. The examining attorney has refused registration under Section 2(d) of the Lanham Act, citing a registration of TIDEWELL for wine owned by Tidewell Cellars, a family winery in Sonoma County. The refusal is a few pages long and attaches a dozen third-party registrations showing that brewers and wineries often sell canned cocktails under the same brand.

Marisol already knows two facts the examining attorney does not. First, Tidewell Brewing and Tidewell Cellars have used their names for nine years without a single misdirected order, call, or review that anyone can find. Second, the winery's founder, Ben Okafor, met Tidewell's chief executive at a trade show three years ago, and the two of them laughed about the name. Nobody wrote anything down.

She has four options. She can argue the refusal on the merits and probably lose, because the goods are related and the marks are identical. She can try to cancel the winery's registration, which would be expensive, uncertain, and the end of a friendly relationship. She can buy the winery's mark, which the winery will not sell. Or she can ask the winery to sign an agreement saying that the two companies can share the name, and file it with the USPTO.

The fourth option is the subject of this article. It is the most common way that trademark conflicts end, and it is one of the least understood.

The short answer

Two businesses can contract to share similar marks. A coexistence agreement allocates who uses what, where, on which goods, and in what form. A consent agreement is a coexistence agreement (or a shorter letter of consent) submitted to the USPTO to overcome a likelihood-of-confusion refusal.
The USPTO usually gives a detailed agreement great weight. The tenth factor in In re E.I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973) is the "market interface" between the parties, including any agreement between them. The Federal Circuit has repeatedly reversed the Trademark Trial and Appeal Board for discounting a thoughtful agreement, most famously in In re Four Seasons Hotels Ltd., 987 F.2d 1565 (Fed. Cir. 1993).
A "naked" consent can fail. A one-paragraph permission with no explanation and no undertakings carries little weight, as In re Mastic Inc., 829 F.2d 1114 (Fed. Cir. 1987) shows. A "clothed" agreement explains why confusion is unlikely and commits the parties to keep it that way.
The deal must fit the register. The Board reads the application and registration as written. A private promise to sell only in New England does not narrow a nationwide application, which is why the consent in In re Bay State Brewing Co. (T.T.A.B. 2016) failed. Geographic splits belong in a concurrent use proceeding under 15 U.S.C. § 1052(d).
The contract is enforceable, within limits. Courts enforce coexistence agreements as contracts and generally evaluate their restraints under the antitrust rule of reason, as in Clorox Co. v. Sterling Winthrop, Inc., 117 F.3d 50 (2d Cir. 1997). An agreement that uses trademarks as cover for dividing product or geographic markets can be per se illegal.
It does not bind the world. The agreement binds its parties and, usually, their successors. It does not bind the USPTO, third parties, or consumers, and it is not a license. If one party is really letting the other use its mark, quality control becomes necessary.
Draft for the long run. The hard questions arrive years later: an acquisition, a new product line, a bankruptcy, an expansion abroad. The agreements that survive are the ones that anticipated them.

What these agreements are, and what they are not

Lawyers use the vocabulary loosely, so it helps to fix it at the start.

A coexistence agreement is a contract between two owners of similar marks that allocates their rights. It typically recites that both parties use their marks, states that they do not believe confusion is likely if each keeps to its lane, defines the lanes (goods, services, territory, channels, presentation), and adds covenants: not to oppose or cancel each other's registrations, to cooperate if confusion arises, and to bind successors. Coexistence agreements resolve demand letters, oppositions, cancellations, and lawsuits. Some are signed before any dispute, when two companies discover each other in a clearance search.

A consent agreement is a coexistence agreement written, at least in part, for the USPTO. Its purpose is to persuade an examining attorney or the Board that the applicant's mark is registrable despite a cited registration. Often the parties sign a long-form agreement with all of the business terms and a short-form consent that is filed publicly, a practice the Board described in Bay State Brewing.

A letter of consent is the minimal version: a signed statement by the registrant that it consents to registration of the applicant's mark. As the cases below show, a letter that says only that is the weakest form of evidence.

A concurrent use registration is different in kind. It is a registration that the USPTO issues with a geographic restriction, under the proviso to Section 2(d), after a concurrent use proceeding or a court decree. A coexistence agreement that splits the country can support a concurrent use application, but the agreement alone does not produce a territorially restricted registration.

A license is different again. In a license, the trademark owner permits another to use the owner's mark, and the owner must control the nature and quality of the goods sold under it. In a coexistence agreement, each party uses its own mark as its own source identifier. Neither is supposed to be using the other's mark, and neither needs to control the other's quality. The distinction matters because a coexistence agreement that is really a license without control can put a mark at risk of abandonment, a problem discussed in Trademark Licensing and Quality Control: How Naked Licensing Kills a Brand.

Instrument	Who signs	What it does	Where it is filed	Main risk
Coexistence agreement	Both mark owners	Allocates fields, territories, presentation; adds no-challenge and cooperation covenants	Usually private	Antitrust exposure if it divides product markets; gaps when businesses change
Consent agreement	Both owners (sometimes only registrant)	Explains to the USPTO why confusion is unlikely	Filed with an office action response or appeal	Rejected as "naked" if conclusory
Letter of consent	Registrant	States consent to registration	Filed with the USPTO	Little weight on its own
Concurrent use registration	Issued by the USPTO	Registers a mark for a defined territory	Concurrent use proceeding	Requires qualifying prior use (or consent) and Board approval
License	Owner and licensee	Permits licensee to use owner's mark	Usually private; may be recorded	Naked licensing and abandonment if no quality control

Why rivals sign them

The plainest reason is cost. A contested opposition or cancellation before the Trademark Trial and Appeal Board can run for two years and cost each side a large share of what a modest federal lawsuit would cost; an infringement action costs more. A coexistence agreement can be negotiated in weeks and costs a small fraction of either.

The second reason is uncertainty. Likelihood of confusion is a multi-factor judgment, and reasonable decision makers reach different results on the same facts. The Federal Circuit applies the thirteen DuPont factors; regional circuits apply their own lists, such as the Polaroid Corp. v. Polarad Electronics Corp., 287 F.2d 492 (2d Cir. 1961) factors in the Second Circuit and the AMF Inc. v. Sleekcraft Boats, 599 F.2d 341 (9th Cir. 1979) factors in the Ninth. A party that cannot predict the outcome often prefers a negotiated line it can see.

The third reason is that a Board decision can follow the parties into court. In B&B Hardware, Inc. v. Hargis Industries, Inc., 575 U.S. 138 (2015), the Supreme Court held that a TTAB likelihood-of-confusion finding can have issue-preclusive effect in later infringement litigation when the usages adjudicated are materially the same. That raised the stakes of losing an opposition, and it made negotiated outcomes more attractive.

The fourth reason is commercial. Two companies that have coexisted quietly for years usually have no interest in a fight. The applicant wants a registration; the registrant wants assurance that the applicant will not drift into its core market and will not attack its registration. Each side can give the other what it wants at little cost to itself. The negotiation techniques that get parties to that point are the same ones used in any litigation settlement, explained in Settlement Negotiation Tactics: Principled Negotiation, Anchoring, and Handling Hardball in Litigation. The trademark-specific process, step by step, is in Negotiating a Trademark Coexistence Agreement: A Practical Guide.

The doctrine: from National Distillers to DuPont

Section 2(d) of the Lanham Act bars registration of a mark that "so resembles" a registered mark, or a mark previously used and not abandoned, "as to be likely, when used on or in connection with the goods of the applicant, to cause confusion, or to cause mistake, or to deceive." The statute does not mention consent. The role of agreements comes from case law, and the case law comes largely from the Court of Customs and Patent Appeals and its successor, the Federal Circuit.

National Distillers and the self-interest of owners

The foundation is a concurring opinion. In In re National Distillers & Chemical Corp., 297 F.2d 941 (C.C.P.A. 1962), the court reversed a refusal where the applicant and the owner of the cited registration had agreed to the registration. Judge Giles Rich, concurring, argued that the Patent Office's job was not to deny registration whenever it independently perceived "any likelihood of confusion of any kind," without regard to the agreements of the owner of the prior registration. His central point, later quoted by the Federal Circuit in Four Seasons, was practical: denying a registration does not stop anyone from using a mark. "No government could police trademark use so as to protect the public from confusion. It must count on the self-interest of trademark owners to do that."

That insight explains everything that follows. A refusal to register does not prevent confusion in the marketplace; it only keeps a mark that will be used anyway off the register. And the businesses most hurt by confusion are the two that are being asked to tolerate each other.

DuPont and factor ten

In DuPont, the court took up a refusal of RALLY for an automobile polish and cleaner in light of an earlier registration of RALLY for a cleaning product. The two companies had entered into an agreement restricting their respective uses of the mark. The Board held that the agreement did not overcome the likelihood of confusion. The C.C.P.A. reversed, and in doing so Judge Markey set out the thirteen factors that now structure every Section 2(d) analysis at the USPTO. The tenth is "the market interface between applicant and the owner of a prior mark," which the court broke into four parts: a mere "consent" to register or use; agreement provisions designed to preclude confusion, such as limitations on continued use of the marks by each party; an assignment of the mark, application, registration, and goodwill of the related business; and laches and estoppel attributable to the owner of the prior mark and indicative of a lack of confusion.

Three passages from DuPont are quoted in nearly every consent case since:

"[W]hen those most familiar with use in the marketplace and most interested in precluding confusion enter agreements designed to avoid it, the scales of evidence are clearly tilted. It is at least difficult to maintain a subjective view that confusion will occur when those directly concerned say it won't."
"Decisions of men who stand to lose if wrong are normally more reliable than those of examiners and judges."
A "naked 'consent' may carry little weight," while the weight given to more detailed agreements "should be substantial."

The DuPont framework remains the law. The Federal Circuit applies it to every Board decision, including inter partes decisions such as Coach Services, Inc. v. Triumph Learning LLC, 668 F.3d 1356 (Fed. Cir. 2012), and the Board must consider each factor for which there is evidence. A consent agreement is therefore evidence, not a trump card, and its weight depends on what it says and what the rest of the record shows. For a full walk through the other factors, see Trademark Likelihood-of-Confusion Analysis: A Practical Checklist and Navigating the Maze of Trademark Confusion: Key Considerations for Brand Owners.

The key cases, told as stories

The doctrine is best learned through the cases, because each one turns on what the agreement said and what the parties were trying to accomplish.

In re N.A.D.: competitors who "clearly thought out their commercial interests"

North American Drager, a Pennsylvania company majority-owned by the German firm Draegerwerk AG, applied to register NARKOMED for anesthesia machines. The examining attorney refused registration over two registrations owned by Narco Scientific: NARCO MEDICAL SERVICES for renting and leasing hospital equipment, and NARCO and design for a long list of medical equipment that included "apparatus for administration of anesthesia." The goods were, in part, identical.

The parties had a history. Draegerwerk had brought cancellation proceedings against Narco's registrations, and the fights had involved several marks over several years. They ended in written settlement agreements in which Draegerwerk and N.A.D. abandoned some applications, phased out four other NARKO- marks, and paid money. Narco, in turn, expressly acknowledged N.A.D.'s right "to the use and registration of the trademark NARKOMED." The examining attorney and the Board refused to give the consent any weight, reasoning that confusion was certain and the parties had not explained how customer confusion would be avoided.

In In re N.A.D. Inc., 754 F.2d 996 (Fed. Cir. 1985), the Federal Circuit reversed. The court emphasized that these were competitors in a restricted field who "clearly thought out their commercial interests with care," that it was "highly unlikely that they would have deliberately created a situation in which the sources of their respective products would be confused by their customers," and that the buyers of anesthesia equipment are sophisticated. It also delivered the line later quoted in Bongrain: "We have never found anything axiomatic about the application of § 2(d) to fact situations, especially when consent agreements are involved."

N.A.D. teaches two things. A consent negotiated at arm's length, with real concessions on both sides, carries more weight than one given for free. And the purchasers' sophistication can do much of the work that explicit marketplace restrictions would otherwise do.

Bongrain: the Board that cancelled registrations nobody asked it to cancel

Delice de France sold croissants, brioche, and other bakery products under a design mark and applied to register it. The application was refused over two registrations of LE PETIT DELICE DE FRANCE for cheese and milk products, owned by Bongrain. The two companies' lawyers talked, studied the market, found no confusion, and signed a four-page agreement in which Delice agreed not to use its mark on cheese or milk products and Bongrain agreed not to use its mark on bakery products. Delice also promised not to oppose or cancel Bongrain's registrations.

The examining attorney was unmoved. To remove the obstacle, Delice petitioned to cancel Bongrain's registrations, and the parties submitted the matter on stipulated facts, with Bongrain moving for summary judgment that confusion was unlikely. Both parties were plainly seeking the same result: registrations for both. The Board instead granted summary judgment for Delice and cancelled Bongrain's registrations.

In Bongrain International (American) Corp. v. Delice de France, Inc., 811 F.2d 1479 (Fed. Cir. 1987), the Federal Circuit reversed in pointed language. Parties to such agreements, it said, "are in a much better position to know the real life situation than bureaucrats or judges and therefore such agreements may, depending on the circumstances, carry great weight." The Board's "independent, misguided efforts" to prove a case of confusion "when not asked to do so" did not serve the policies of the Act, one of which is to put on the register as many of the marks in actual use as possible so that they are available for search.

The practical lesson of Bongrain is about structure. When a consent is rejected ex parte, an inter partes proceeding on stipulated facts can put the question back before the Board, but it carries risk: the Board may do something neither party wants. Agreements should anticipate that possibility, for example by making the no-challenge covenant survive and by requiring the parties to cooperate in any appeal.

In re Mastic: eleven factors, no facts

Mastic Inc., a Canadian company, applied under Section 44 to register SHURLOK for vinyl siding, based on its Canadian registration and without alleging any use in the United States. The examining attorney cited The Flintkote Company's registrations of SHURLOK for roofing shingles. Mastic filed a consent signed on behalf of Flintkote. The consent recited that the parties had examined their products and marketing and concluded there was no likelihood of confusion, and it listed eleven factors the parties had taken into account, tracking the DuPont factors almost word for word: dissimilar trade channels, different buyers, no actual confusion, and so on. The examining attorney put in advertisements showing that building supply companies and contractors handle both siding and roofing.

The Federal Circuit affirmed the refusal. Its reasoning is the most useful passage in the case law for anyone drafting a consent:

DuPont does not make it a "given" that businesses consent only when confusion is unlikely. "One must look at all of the surrounding circumstances," the court said, to determine whether the consent reflects marketplace reality "or if the parties struck a bargain that may be beneficial to their own interests, regardless of confusion of the public."
Even a naked consent is "significant additional evidence" if the record otherwise supports the applicant. A consent "clothed" with the parties' agreement to undertake specific arrangements to avoid confusion is entitled to greater weight, "not because of the consent itself, but because such arrangements are additional factors."
The Flintkote document consented only to registration, not use, and Mastic was not using the mark in the United States at all. A Section 2(d) determination assumes both marks are exposed to the public.
The document was a unilateral statement, not a bilateral agreement, and any arrangement about trade channels "has not been disclosed."
Mechanically naming the DuPont factors does not give a consent "greater dignity." A statement that the parties considered the "dissimilarity of the established, likely to continue, trade channels" does not prove that the channels are dissimilar. "Evidence must be submitted."

The court added a point that every drafter should remember: if the goods "are likely to be attributed to the same source because of the use of the same or a similar mark, a license (not merely a consent) is necessary to cure the conflict."

In re Four Seasons Hotels: "a cat watching the wrong rat hole"

The Santa Barbara Biltmore, an oceanfront resort open since the 1920s, came under Four Seasons management in 1987, and Four Seasons applied to register FOUR SEASONS BILTMORE for resort innkeeping services. The examining attorney cited THE BILTMORE LOS ANGELES for hotel services, owned by T.A.T. Los Angeles Co., a downtown hotel 120 miles to the south that had also operated since the 1920s. The two owners signed a consent agreement that did several specific things. Four Seasons would use and register BILTMORE only for the Santa Barbara resort, and only as FOUR SEASONS BILTMORE or SANTA BARBARA BILTMORE. It would not use BILTMORE for any other hotel. T.A.T. would not use FOUR SEASONS. Neither would attempt to associate itself with the other. Four Seasons could advertise worldwide. And both would cooperate to eliminate or minimize confusion if any arose.

The Board treated the agreement as a concurrent use agreement, gave it little weight, and affirmed the refusal. The Federal Circuit reversed, in In re Four Seasons Hotels Ltd., 987 F.2d 1565 (Fed. Cir. 1993). Quoting a colleague, the court said the PTO "is, at times, like a cat watching the wrong rat hole." Its role is not to second-guess "the conclusions of those most familiar with the marketplace." The services were similar but not the same: a resort with outdoor activities at the ocean versus a traditional city hotel. The marks shared an element, but a common element does not compel a finding of confusion. And because the agreement "constitute[d] far more than mere 'consent,'" it deserved to play a "dominant role."

Four Seasons is the model for a clothed consent. It names the exact form of the mark each party will use, ties use to a specific property, adds reciprocal non-association covenants, and includes a cooperation clause. It also reflects a real marketplace history: seventy years of two hotels with the same word in their names.

Amalgamated Bank: a negotiated geographic line

Two "labor banks" had used AMALGAMATED for decades, one in New York since 1923 and one in Chicago since 1922. Each advertised nationally and had clients across the country, but neither had a branch in the other's state. When the New York bank applied to register AMALGAMATED and AMALGAMATED BANK, the Chicago bank sought extensions to oppose and then signed a settlement agreement: it would not oppose and would take no action against the New York bank's use throughout the United States except Illinois, and the New York bank would not interfere with the Chicago bank's continued use. The Board denied the New York bank a concurrent registration. In Amalgamated Bank of New York v. Amalgamated Trust & Savings Bank, 842 F.2d 1270 (Fed. Cir. 1988), the Federal Circuit reversed and remanded, criticizing the Board for presuming that its understanding of the banking business exceeded that of the banks themselves.

In re Bay State Brewing: when the deal does not match the application

The modern cautionary tale is a beer case. Bay State Brewing applied to register TIME TRAVELER BLONDE for beer. The examining attorney cited TIME TRAVELER for "beer, ale and lager." Bay State conceded on appeal that the marks were similar and the goods related, and argued only that it had a consent agreement with the registrant. In its own brief, it said that "the parties acknowledge that confusion is likely unless they both adhere to the terms" of the agreement.

The parties had signed a long-form agreement and a short-form agreement intended for the public record. The key restriction was geographic: Bay State would use its mark only in New England and New York. The registrant's use was not limited at all.

The Board affirmed the refusal in In re Bay State Brewing Co. (T.T.A.B. Feb. 25, 2016), a precedential decision. It identified two problems. First, the agreement did not separate the parties at all inside New England and New York, because the registrant remained free to sell there; both marks would be on the same shelves in the same states. Second, the geographic restriction was not in the application and would not appear in any registration. A nationwide registration would give Bay State presumptive nationwide exclusive rights under 15 U.S.C. § 1057(b), and the register "should reflect, within the constraints of the Trademark Act, the realities of the marketplace." With virtually identical marks, identical goods and channels, and impulse purchases, the agreement was outweighed.

Bay State is often described as a case in which the Board "rejected a consent agreement." It is more precise to say that the Board rejected this agreement, because it did not do the work an agreement must do. A geographic split needed a concurrent use application. An agreement that conceded confusion absent its terms, while allowing overlap in the same territory, could not persuade anyone.

In re American Cruise Lines: identical services, industry practice, and two consents

Two years later the Board reached the opposite result on facts that look, at first glance, just as hard. American Cruise Lines applied to register AMERICAN CONSTELLATION for cruise ship services. The examining attorney cited CONSTELLATION and CELEBRITY CONSTELLATION, both registered to the same owner for cruise ship services. The services were identical.

In In re American Cruise Lines, Inc. (T.T.A.B. Oct. 3, 2018), the Board reversed. It relied on evidence that cruise purchasers exercise a heightened degree of care, on the industry practice of different cruise lines using some of the same terms to name ships without confusion, and on two consents to use and register executed by the applicant and the registrant. The Board restated the governing rule: "clothed" consent agreements, where competitors have clearly thought out their commercial interests, should be given great weight, and the USPTO should not substitute its judgment for that of the real parties in interest without good reason, that is, unless the other relevant factors clearly dictate a finding of likelihood of confusion.

The contrast with Bay State is instructive. Both involved nearly identical marks and identical goods or services. In American Cruise Lines, the record contained independent evidence (careful purchasers, an industry naming convention) that supported the parties' judgment. In Bay State, the record contained the applicant's own concession that confusion was likely without the agreement, and an agreement that did not actually separate the marks.

Wacker Neuson and Dare Foods: relationship and track record

Two other precedential Board decisions round out the picture. In In re Wacker Neuson SE (T.T.A.B. Dec. 28, 2010), the applicant sought to register WACKER NEUSON for construction machinery over a Supplemental Register registration of NEUSON for excavators. The consent documents were thin, but the Board found that the relationship between the parties, which belonged to an overarching corporate group, made the consent viable. It added that a registrant's consent removes the basis for the usual rule that doubts are resolved in favor of the registrant, citing In re Donnay International, 31 USPQ2d 1953 (T.T.A.B. 1994).

In In re Dare Foods Inc. (T.T.A.B. Mar. 29, 2022), the Board reversed a refusal of RAINCOAST DIP for snack food dips over RAINCOAST TRADING for seafood products. The coexistence agreement, which the Board reproduced as an appendix, recited the parties' differences, committed them to cooperate if actual confusion arose, and contained mutual promises not to oppose, cancel, or limit each other's marks. The applicant pointed out that the arrangement had governed the parties' marks for nearly eight years. Reproducing the agreement in the opinion gives practitioners a useful example of what a successful clothed agreement looks like.

Naked versus clothed: what makes an agreement persuasive

Read together, the cases produce a working test. The Trademark Manual of Examining Procedure states the same principles in its section on consent agreements (TMEP § 1207.01(d)(viii)): examining attorneys should give substantial weight to a proper clothed agreement and should not substitute their judgment for that of the parties absent other factors that clearly point to confusion, while a naked consent is generally given little weight. The USPTO revises the TMEP periodically, so check the current edition for the exact language.

Feature	Naked consent	Clothed agreement
Signatories	Registrant only	Both parties, by authorized officers
Scope	Consent to registration	Consent to use and registration
Reasons	None, or a list of DuPont factors	Specific facts: different goods, buyers, channels, presentation; history of coexistence; absence of actual confusion with time frame
Undertakings	None	Restrictions on fields, forms of the mark, trade dress, channels, or territory; non-association covenants
Cooperation	None	Duty to notify and to act to remedy any actual confusion
Fit with the register	Not addressed	Identifications amended where needed so the register reflects the restrictions
Concessions	None	Recital that confusion is unlikely (not that it is likely "unless" the agreement is honored)

Three drafting points come up repeatedly.

State facts, not conclusions. Mastic is direct authority that reciting the DuPont factors proves nothing. The agreement should say what the channels actually are ("Applicant sells only through licensed liquor retailers in [STATES]; Registrant sells only through its tasting room and wine club"), what the goods actually are, who buys them, and for how long the parties have coexisted without known confusion.

Avoid admissions. Do not write that confusion would be likely but for the agreement. Bay State turned in part on that concession. A better formulation states that the parties believe confusion is unlikely given the differences between their goods and marks, and that they have nonetheless agreed to additional measures to keep it that way.

Make the register match the deal. The Board decides registrability on the identifications in the application and the cited registration, not on how the parties actually do business. Octocom Systems, Inc. v. Houston Computer Services, Inc., 918 F.2d 937 (Fed. Cir. 1990) and Stone Lion Capital Partners, L.P. v. Lion Capital LLP, 746 F.3d 1317 (Fed. Cir. 2014) hold that the analysis turns on the goods or services as identified, and In re Shell Oil Co., 992 F.2d 1204 (Fed. Cir. 1993) applied the same principle to unrestricted identifications. Conversely, a real restriction written into an identification counts: in In re i.am.symbolic, llc, 866 F.3d 1315 (Fed. Cir. 2017), the court considered a restriction in the identification but held that it did not avoid confusion with the cited marks. If the agreement limits the applicant to snack dips sold in grocery stores, amend the identification to say so where the rules permit, or ask the registrant to narrow its own registration under 15 U.S.C. § 1068 or by a voluntary amendment. The same logic applies to presentation: a standard character registration covers the word in any font, as In re Viterra Inc., 671 F.3d 1358 (Fed. Cir. 2012) holds, so a promise to use a different logo does little at the USPTO if both parties hold or seek standard character registrations.

The mechanics of responding to a refusal with a consent, and the evidence that should accompany it, are covered in Responding to a Trademark Office Action Checklist: A Practical Checklist, and the full drafting checklist for both the private and public documents is Trademark Coexistence and Consent Agreement Checklist: A Practical Checklist.

Geographic splits and concurrent use registration

Many real coexistence stories are geographic. Two restaurants, two banks, or two hardware stores adopted the same name in different cities before either knew of the other. The common law has handled this for more than a century. Under Hanover Star Milling Co. v. Metcalf, 240 U.S. 403 (1916) and United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90 (1918), a good-faith remote user can acquire rights in its own territory even against an earlier user elsewhere. That doctrine, and the effect of federal registration on it, is explained in Tea Rose-Rectanus and the Geographic Scope of Common-Law Trademark Rights: A Practical Guide.

The Lanham Act accommodates these situations through the proviso to Section 2(d). The USPTO may issue concurrent registrations, with "conditions and limitations as to the mode or place of use," to persons entitled to use the same or similar marks as a result of concurrent lawful use in commerce, where confusion is not likely under those conditions. The concurrent lawful use generally must have begun before the earliest filing date of the applications or registrations involved. The statute relaxes that requirement in one important situation: prior use is not required "when the owner of such application or registration consents to the grant of a concurrent use registration to the applicant." A court that has finally determined the parties' rights may also order concurrent registrations. The procedures appear in the rules of practice at 37 C.F.R. Part 2 (principally § 2.99), and the Board's authority to restrict registrations comes from 15 U.S.C. § 1068.

The leading case on how agreements fit into concurrent use is In re Beatrice Foods Co., 429 F.2d 466 (C.C.P.A. 1970). Beatrice had used HOMESTEAD for margarine since 1953; Fairway Foods, a Minnesota wholesaler, had used HOMESTEAD for dairy products since 1956 and applied first. Beatrice opposed, and then the parties settled by dividing the country into two territories. They asked the Office to issue concurrent registrations matching their agreement. A Board member wrote that adverse parties "may not by agreement resolve a question which this Office has the responsibility to determine" and required proofs. On appeal, the C.C.P.A. held, among other things, that the territorial restrictions must be determined on the facts as they exist when registrability is decided, not frozen at the filing date, and that "any attempt, by the parties themselves, to solve the problem of public confusion, should be given serious consideration." The CCPA's later decision in Weiner King, Inc. v. Wiener King Corp., 615 F.2d 512 (C.C.P.A. 1980), applied the Beatrice framework to two hot dog restaurant chains, one in New Jersey and one based in North Carolina, and Gray v. Daffy Dan's Bargaintown, 823 F.2d 522 (Fed. Cir. 1987) shows the limit: a junior user who admitted using DAFFY DAN'S in the senior user's own New Jersey trading area could not obtain a concurrent use registration, because the overlapping use made confusion likely. Separation has to be real.

The practical rule is short. If the deal is "you take the East and I take the West," the parties need a concurrent use application (and, usually, an amendment of the existing registration) rather than an ordinary consent. Bay State is what happens when parties try to use a consent to accomplish a territorial split. Four Seasons shows the alternative: an agreement that ties one party's use to a single named property, with worldwide advertising permitted, can support an unrestricted registration because the restriction is about the nature of the services and the form of the mark, not a line on a map. For a deeper treatment of territorial rights, see The Geographic Scope of Common-Law Trademark Rights: A Practical Guide.

Enforcing the agreement between the parties

A coexistence agreement is a contract, governed by state contract law and enforced in the ordinary way. Disputes usually arise years after signing, when a business has changed in a way the drafters did not foresee.

The contract governs, not residual common-law rights

Times Mirror Magazines, Inc. v. Field & Stream Licenses Co., 294 F.3d 383 (2d Cir. 2002) is the classic example. Gordon & Ferguson had sold clothing under FIELD & STREAM since 1915 and registered it for clothing in 1926; the magazine had been published since 1895. In 1984, the magazine's then-owner and Gordon & Ferguson signed an agreement: Gordon & Ferguson had exclusive worldwide rights to the mark for "apparel," defined in detail to exclude jewelry and footwear, and the magazine had exclusive rights for publications and for hunting- and fishing-related products it had already sold under the mark, with a six-month window to add to the list. Later agreements adjusted the lines. When the licensing company expanded into new products, the magazine's new owner sued for breach, infringement, and rescission, arguing it retained a residual common-law right in all hunting and fishing goods. The Second Circuit affirmed judgment for the licensing company: "no such residual right exists and . . . the parties' relative rights to the mark are governed by their agreements." The district court had also declined to rescind the contracts in the name of the public interest, finding that the confusion the agreements allowed would not harm consumers.

Two lessons follow. The definitions of fields are the agreement. If "apparel" is defined, every unlisted product category will be argued about, so the drafters must decide what happens to new categories: reserved to one party, open to both with notice, or subject to a consultation process. And a party that keeps doing business under an agreement after learning of a breach risks waiver, as the magazine discovered.

No-challenge covenants are generally enforced

Most coexistence agreements contain a promise not to oppose, petition to cancel, or otherwise attack the other party's registrations. Patent law has long been skeptical of such clauses after Lear, Inc. v. Adkins, 395 U.S. 653 (1969), which held that a patent licensee may challenge validity despite a contractual promise not to. Trademark courts have generally declined to extend Lear. In Beer Nuts, Inc. v. King Nut Co., 477 F.2d 326 (6th Cir. 1973), King Nut had settled an earlier dispute over KING'S BREW NUTS by promising to stop using the name and recognizing the validity of BEER NUTS, in a contract stated to be perpetual and binding on assigns. A decade later, sued over a new package, King Nut argued that BEER NUTS had become descriptive. The Sixth Circuit held that the settlement barred the defense, reasoning that the public interest in the trademark context "is not so great that it should take precedence over the rule of the law of contracts that a person should be held to his undertakings." The Board likewise recognizes contractual estoppel as a defense in inter partes proceedings, as described in the TTAB Manual of Procedure. Drafters should still write the covenant carefully: some courts weigh the public interest more heavily where a mark has become generic, and a covenant that bars a challenge on every possible ground forever invites argument.

Releases cover what they say, and no more

The parties' release is the other half of the bargain, and its scope controls later suits. In Lucky Brand Dungarees, Inc. v. Marcel Fashions Group, Inc., 590 U.S. 405 (2020), a 2003 settlement had ended a dispute over GET LUCKY: Lucky Brand agreed to stop using the phrase, and Marcel released claims arising out of Lucky Brand's use of its own marks. Years of further litigation followed. When Marcel sued again, over later conduct, Lucky Brand raised the release. The Supreme Court held that Lucky Brand was not precluded from raising that defense merely because it had not pressed it in an earlier round, since the later suit challenged different conduct. The decision is a reminder that every round of trademark litigation turns on what the settlement actually released and when. General settlement drafting is covered in Drafting a Settlement Agreement and Release: A Practical Guide and Settlement Agreements, Releases, and Rule 68: Ending a Case on Terms That Hold.

Tolerance is not an agreement

Some coexistence is informal: years of mutual awareness and silence. That is not the same as a contract, and it does not give the junior user the same protection. In Kellogg Co. v. Exxon Corp., 209 F.3d 562 (6th Cir. 2000), Kellogg's Tony the Tiger and Exxon's cartoon tiger had coexisted for decades, and in 1968 Kellogg had even asked Exxon not to oppose Kellogg's registration in Germany. When Exxon began using its tiger to promote food and beverages at its convenience stores, Kellogg sued. The district court held that Kellogg had acquiesced; the Sixth Circuit reversed, holding that the acquiescence finding was error and that fact questions remained. A written agreement that defined the fields would have avoided years of litigation about what each side had implicitly accepted. Equitable defenses in inter partes proceedings are available under 15 U.S.C. § 1069, but they are a poor substitute for a contract.

Remedies

Breach of a coexistence agreement usually supports both a contract claim and, where the breaching use is outside the permitted field, a claim for infringement under 15 U.S.C. § 1114 or § 1125(a). Injunctive relief is available under § 1116, and monetary relief under § 1117. Many agreements add a stipulation that breach causes irreparable harm, a notice-and-cure period, a consent to specific performance, and a dispute resolution clause. Courts treat stipulated irreparable harm as relevant but not binding when deciding on preliminary relief under Fed. R. Civ. P. 65. If the parties choose arbitration, the Federal Arbitration Act, 9 U.S.C. § 2, makes the clause enforceable, but an arbitrator cannot cancel a federal registration; the agreement should require the losing party to file whatever USPTO papers the award requires.

Antitrust: where coexistence becomes market division

Coexistence agreements are agreements between competitors, or at least potential competitors, about where and how they will sell. Section 1 of the Sherman Act, 15 U.S.C. § 1, prohibits contracts in unreasonable restraint of trade. The question is which standard applies.

The danger case is Timken Roller Bearing Co. v. United States, 341 U.S. 593 (1951). American, British, and French bearing makers had divided world markets and fixed prices, and they defended the arrangements as trademark licensing. The Supreme Court rejected the defense: trademark agreements cannot be used as a cover for a territorial cartel. Horizontal territorial allocation is per se unlawful, as the Court confirmed in Palmer v. BRG of Georgia, Inc., 498 U.S. 46 (1990).

The safe-harbor case is Clorox. The owners of PINE-SOL and LYSOL had fought since the 1950s and signed agreements in 1967 and 1987. The 1987 agreement regulated how PINE-SOL disinfectant products could be advertised and packaged and restricted the types of PINE-SOL disinfectant products that could be sold. After Clorox acquired PINE-SOL, it sued under Sections 1 and 2 of the Sherman Act, arguing that the restrictions had no remaining trademark purpose and kept the popular brand out of markets LYSOL dominated. The Second Circuit affirmed summary judgment for the LYSOL owners. Unlike "trademark agreements that in reality serve to divide markets," as in Timken, the agreement "merely regulates the way a competitor can use a competing mark," so the rule of reason applied, as it has since Board of Trade of Chicago v. United States, 246 U.S. 231 (1918). Clorox could not show harm to competition: large companies could enter the markets with other brands, and Clorox itself remained free to compete with PINE-SOL in most respects. The court added that "trademark agreements are favored in the law" as a means of reducing confusion and avoiding litigation, and that absent evidence that the trademark provisions are auxiliary to an underlying illegal agreement, "the parties' determination of the scope of needed trademark protections is entitled to substantial weight."

The modern application is 1-800 Contacts, Inc. v. FTC, 1 F.4th 102 (2d Cir. 2021). 1-800 Contacts had settled trademark disputes with rival online contact lens sellers through agreements in which the parties agreed not to bid on each other's trademarks as search keywords and to use negative keywords. The FTC held the agreements unlawful. The Second Circuit vacated the order, applying the rule of reason and holding that trademark settlement agreements are not inherently suspect; the procompetitive justifications of protecting trademarks and avoiding litigation had to be weighed. The case also shows that regulators do look at trademark settlements, as the Supreme Court made clear about patent settlements in FTC v. Actavis, Inc., 570 U.S. 136 (2013).

The practical lines are these. Restrictions on the form, presentation, and field of use of a mark, reasonably tied to avoiding confusion, are evaluated under the rule of reason and usually survive. Restrictions on what products a party may sell under any name, what prices it may charge, which customers it may serve, or where it may sell (as opposed to where it may use a particular mark) look like market division. An agreement should say that each party remains free to sell any product anywhere under a different mark, and the restrictions should be no broader than the confusion problem requires. These issues are discussed more broadly in Antitrust for Technology Companies: Monopolization, Platform Conduct, and the IP-Antitrust Interface and Antitrust Compliance for Distribution and Pricing: Resale Price Maintenance, Colgate, and Territory Restrictions.

Outside the United States, competition authorities apply their own rules. In the European Union, the Court of Justice held in BAT Cigaretten-Fabriken GmbH v. Commission (Case 35/83, [1985] ECR 363), the Toltecs/Dorcet case, that a trademark delimitation agreement can infringe the EU competition rules when it goes beyond what is needed to avoid confusion. Global agreements should be reviewed by local counsel in each major jurisdiction.

Naked licensing risk: when coexistence becomes a license

A coexistence agreement assumes two sources. Each party uses its own mark for its own goods, and the public is expected to distinguish them. Trouble starts when the agreement, or the parties' practice, implies one source: for example, when one party is permitted to use the other's exact mark, or when the parties begin co-branding, cross-selling, or describing themselves as affiliated.

At that point the arrangement starts to look like a license, and a license without quality control is a "naked license" that can cause abandonment. The Lanham Act defines an abandoned mark to include one that, through the owner's course of conduct, loses its significance as a mark (15 U.S.C. § 1127), and treats use by "related companies" controlled by the owner as the owner's use (15 U.S.C. § 1055). The duty to control dates to Dawn Donut Co. v. Hart's Food Stores, Inc., 267 F.2d 358 (2d Cir. 1959). Courts have found naked licensing and abandonment in Barcamerica International USA Trust v. Tyfield Importers, Inc., 289 F.3d 589 (9th Cir. 2002) and FreecycleSunnyvale v. Freecycle Network, 626 F.3d 509 (9th Cir. 2010), and the Seventh Circuit applied the doctrine in Eva's Bridal Ltd. v. Halanick Enterprises, Inc., 639 F.3d 788 (7th Cir. 2011).

Three rules keep a coexistence agreement on the right side of the line. First, do not grant the other party the right to use your mark; recognize its right to use its own. Second, require visible differentiation (house marks, distinct trade dress, different forms of the shared word) so that the two sources remain distinct. Third, if a license is genuinely part of the deal (for example, a license-back for a narrow product where the fields overlap), write it as a license with quality control, inspection, and approval rights, using the provisions described in Drafting and Administering a Trademark License: A Practical Guide. Where the parties are members of one corporate group, the related-company rules of Section 5 and decisions such as In re Wella A.G., 787 F.2d 1549 (Fed. Cir. 1986) may matter more than consent, a point echoed in Wacker Neuson.

Effect on third parties and on later disputes

A coexistence agreement binds only its parties and those who succeed to their rights. It has four limits that clients often misunderstand.

It does not bind the USPTO. The examining attorney and the Board weigh the agreement as evidence. Mastic and Bay State are proof that they can reject it.

It does not bind third parties. A competitor who opposes the applicant's mark on its own Section 2(d) ground is not affected by the consent of a different registrant. A third party's own registration can still block the application.

It does not change the marketplace. If actual confusion develops, the agreement does not make it disappear. A court in later litigation between the parties will usually treat the agreement as strong evidence of what the parties thought, but it can still find confusion, particularly where one party has moved outside its lane. Practitioners sometimes worry that a consent will be used against them in litigation with strangers, for example as evidence that the shared term is weak because two owners use it. That argument is available to a defendant, and it is one reason registrants prefer agreements that stress the differences between the two marks rather than conceding that the shared term is descriptive.

It may shape preclusion. A registration obtained with a consent is a registration like any other, with the evidentiary presumptions of 15 U.S.C. § 1115 and the possibility of incontestability under § 1065, and, under B&B Hardware, contested Board findings can bind later. Parties settling an opposition should consider whether to dismiss with or without prejudice and what the dismissal will mean in later litigation. Dismissal mechanics in federal court are governed by Fed. R. Civ. P. 41; the Board's own rules appear in 37 C.F.R. Part 2 and the TTAB Manual of Procedure. The TTAB settlement path is covered in the TTAB Practice Toolkit: Oppositions, Cancellations, and the Trademark Trial and Appeal Board from Pleading to Appeal.

Assignment, successors, and bankruptcy

Coexistence agreements last a long time, often longer than the companies that sign them. Clorox involved an agreement signed by predecessors twenty years before the lawsuit and enforced against the company that bought the brand. Beer Nuts enforced a promise stated to bind "assigns." Times Mirror enforced agreements against a new owner of the magazine.

Trademarks are assigned with their goodwill under 15 U.S.C. § 1060, and a buyer of a mark generally takes it subject to the burdens that run with it. The agreement should nonetheless say so expressly: that it binds and benefits successors and assigns of the marks, that neither party may assign its rights under the agreement except together with the marks and goodwill, and that any assignee must agree in writing to be bound. Change-of-control provisions deserve attention too. A small winery that coexists happily with a small brewery may feel very differently when the brewery is acquired by a national beverage company with a portfolio of ready-to-drink products. Some agreements give the non-transferring party a right to renegotiate specified terms on a change of control; most simply rely on the field restrictions to contain the new owner. Due diligence on acquisitions should always ask for coexistence agreements, a point made in the IP Due Diligence Checklist for Mergers and Acquisitions: A Practical Checklist. Recording the agreement against the affected registrations is optional, but recording any accompanying assignment is not, as discussed in Recording a Trademark Assignment: A Practical Checklist.

Bankruptcy is the other stress test. A coexistence agreement with continuing obligations on both sides (stay out of each other's fields, cooperate, do not challenge) is likely an executory contract under 11 U.S.C. § 365, which a debtor may assume or reject. Section 365(n), which protects licensees of "intellectual property," does not help directly, because the definition in 11 U.S.C. § 101 omits trademarks. The Supreme Court's decision in Mission Product Holdings, Inc. v. Tempnology, LLC, 587 U.S. 370 (2019) supplies the key rule: rejection of an executory contract is a breach, not a rescission, so it does not revoke rights the counterparty already holds under the contract. For a trademark licensee, that meant the debtor's rejection did not terminate the license. For a coexistence counterparty, it suggests that a debtor's rejection would leave the counterparty with a damages claim for the debtor's future non-performance while not stripping the counterparty of the rights it had been granted, although the application to coexistence covenants (as opposed to licenses) has not been squarely litigated. Some courts have also held that an agreement is not executory at all once one side has substantially performed, as in In re Exide Technologies, 607 F.3d 957 (3d Cir. 2010). A buyer of a debtor's brand in a sale under 11 U.S.C. § 363 will want to know whether the coexistence burdens come with it. The bankruptcy rules are explained in Intellectual Property Licenses in Bankruptcy: Section 365(n), Mission Product, and Protecting Your License.

Global coexistence

Trademark rights are territorial, and the Supreme Court reaffirmed that principle for the Lanham Act in Abitron Austria GmbH v. Hetronic International, Inc., 600 U.S. 412 (2023), which limited the Act's infringement provisions to infringing "use in commerce" that is domestic. A party that wants protection abroad needs foreign rights and, often, foreign agreements. The territorial framework is explained in Protecting a Trademark Abroad: The Madrid Protocol, Territoriality, and Abitron v. Hetronic.

Worldwide coexistence agreements are common among multinational companies. They raise distinct issues: which country's law governs, whether a single agreement can be filed with multiple trademark offices (many offices accept letters of consent, some do not, and some give them little weight), whether local competition law restricts territorial allocations, and how to handle countries where neither party has yet filed. The history of the BAYER mark illustrates how long these arrangements can last. The American rights were seized from the German company during World War I and sold to Sterling Drug, which led to decades of disputes and agreements. In Sterling Drug, Inc. v. Bayer AG, 14 F.3d 733 (2d Cir. 1994), the Second Circuit upheld Sterling's contract and infringement claims but vacated an injunction that broadly barred Bayer AG from using its name in materials likely to reach the United States, holding it overbroad as to conduct abroad. A global agreement should anticipate cross-border advertising, websites, and social media, which no longer respect national lines.

The best-known global coexistence story involves two companies called Apple. As recounted in the English High Court's judgment in Apple Corps Ltd v. Apple Computer, Inc., [2006] EWHC 996 (Ch), and widely in the press, Apple Corps (the Beatles' company) and Apple Computer signed a coexistence agreement in 1981 after a dispute over the name, and a second agreement in 1991 after Apple Computer added music capabilities to its products. In 2006, the English High Court held that Apple Computer's use of its mark for the iTunes Music Store fell within its permitted field under the 1991 agreement. In 2007, the companies announced a new arrangement under which Apple Inc. would own the APPLE marks and license certain of them back to Apple Corps. The saga illustrates three recurring problems: fields defined by technology age badly, an agreement that seemed clear in 1991 produced litigation fifteen years later, and the cleanest resolution was a change of ownership with a license, not another round of field definitions.

Tax and accounting

Money often changes hands in a coexistence deal, as it did in N.A.D. The tax character of a payment depends on what it buys. A payment for a release of past infringement claims, a payment for a covenant not to challenge, a payment for the other party's agreement to limit its own use, and a payment for an actual transfer of trademark rights may be treated differently, and provisions of the Internal Revenue Code dealing with intangibles and with transfers of trademarks and trade names (including 26 U.S.C. § 197 and 26 U.S.C. § 1253) can come into play. The allocation recited in the agreement does not bind the taxing authorities but is evidence of the parties' intent. Clients making or receiving significant payments should get tax advice before the allocation is fixed.

Splits, tensions, and open questions

The law here is stable at its core but contested at the edges.

How much deference when the goods are identical? DuPont, N.A.D., and American Cruise Lines accepted agreements involving identical or overlapping goods or services. Bay State rejected one. The difference lies in the rest of the record (sophisticated buyers, industry naming practice, and actual separation versus overlap), but examining attorneys still vary in how readily they accept consents for identical goods and nearly identical marks. Applicants should assume they need independent evidence, not just the agreement.

Can the parties' own concession defeat their agreement? Bay State treated a concession that confusion was likely "unless" the parties followed their agreement as damaging. The better reading is that a concession is evidence like any other, and that agreements written in terms of affirmative differences fare better than agreements written as confusion-management plans.

What standard governs antitrust review? Clorox and 1-800 Contacts both applied the rule of reason, but neither excuses agreements that divide product or geographic markets. How far a geographic split of trademark use (as opposed to sales) can go before it looks like Timken has not been fully worked out, and the answer may differ between the antitrust agencies and the courts.

Are no-challenge covenants always enforceable? Beer Nuts enforced one, and the Board generally applies contractual estoppel, but the public interest in removing generic terms from the register has led some tribunals to look harder at covenants that would preserve a generic registration.

How do coexistence covenants fare in bankruptcy? Mission Product addressed a license. Whether a debtor that rejects a coexistence agreement can then expand into the counterparty's field, and whether the counterparty can enforce the field restriction against the debtor or a buyer, remains largely untested.

Practical implications for each side

For the applicant facing a refusal. Look at the cited registrant before arguing. If its business is genuinely different, ask for a clothed agreement. Prepare a draft yourself, keep it factual, and propose identification amendments that put the key restrictions on the register. Submit evidence alongside the agreement: declarations about trade channels and customers, the length of coexistence, and the absence of known confusion. The USPTO response deadline does not stop while a registrant's lawyers review a draft, so ask for an extension early. If a refusal stands, the Board appeal is the place for the full record, and an inter partes route like Bongrain is available but riskier.

For the registrant asked to consent. Consent is a business asset. You can give it for free, trade it for restrictions on the applicant, or decline. Ask for what protects your core: a field line, presentation requirements, a promise not to expand toward your goods, a no-challenge covenant, notice of any actual confusion, and a successor clause. Avoid admissions that your mark is weak or descriptive. Consider whether the applicant's registration will later be cited against your own expansion, and reserve the fields you plan to enter.

For parties settling litigation or a TTAB proceeding. A coexistence agreement is usually the substantive core of the settlement, with the release, dismissal, and filings as the procedural shell. Coordinate the agreement with the dismissal terms, the USPTO filings (amendments, withdrawal of oppositions, consents), and the timing of each. The broader negotiation process is in Planning and Running a Litigation Settlement Negotiation: A Practical Guide.

For in-house counsel managing a portfolio. Keep a register of every coexistence agreement, its fields, territories, notice obligations, and renewal or termination events. Review it before every product launch, rebrand, acquisition, and international filing. Product teams rarely know that a forty-year-old agreement forbids a particular product category. The rebrand context, where old agreements frequently surface, is covered in Rebranding Without Regret: Clearance, Transition, and Protecting the Old Mark While Launching the New One.

Extended worked example: Tidewell Brewing and Tidewell Cellars

Return to Marisol Vega and the TIDEWELL refusal. What follows is invented, but every step reflects the doctrine above.

Week 1: Assessing the refusal. Marisol reads the office action and the cited registration. The winery's registration covers "wine" in Class 33, without restriction. Tidewell's application covers "canned cocktails; prepared alcoholic cocktails" in Class 33. The examining attorney's evidence shows that many wineries and breweries sell canned cocktails. On the merits, Marisol rates the chance of overcoming the refusal by argument alone as poor: identical marks, related goods, unrestricted identifications that the Board will presume travel in all normal channels.

She then pulls the facts. Tidewell Brewing sells only in Maine, New Hampshire, and Massachusetts, through distributors to bars and package stores, and its cans carry a lighthouse logo and the phrase "Brewed in Portland, Maine." Tidewell Cellars sells about 4,000 cases a year, mostly direct to consumers through its tasting room and wine club, with labels showing a Sonoma vineyard and the full name TIDEWELL CELLARS. Neither company has a record of misdirected communications. The winery has never sold a canned product.

Week 2: The call. Tidewell's chief executive calls Ben Okafor. The message is simple: we have coexisted for nine years; we would like to keep doing so; we want to register for canned cocktails, and we are willing to commit in writing to things that protect your business. Ben is receptive but cautious. His daughter, who now runs marketing, has been talking about a canned rosé spritzer. He calls the winery's trademark lawyer, Daniel Ruiz.

Weeks 3 to 6: Negotiation. Daniel's opening list is broad: Tidewell must never use TIDEWELL on any wine or wine-based product; must always use its lighthouse logo; must not sell in California; and must agree that the winery may register TIDEWELL for canned wine products. Marisol accepts the first point immediately (Tidewell has no interest in wine), accepts a modified version of the second (TIDEWELL must appear with the word BREWING or with the lighthouse logo on every product and in advertising), and rejects the third. A promise not to sell in California would be a restriction on where Tidewell may sell, which raises the market-division concerns of Timken and Palmer, and it would also be the kind of geographic term that, as in Bay State, does not appear on the register and so does not help with the USPTO. She proposes instead a channel commitment for the cocktails: they will be sold only through licensed distributors and retailers, never through wine clubs or tasting rooms. On the fourth point, the parties agree on a field line: wine and wine-based beverages (including canned wine and wine spritzers) belong to the winery; beer, malt beverages, spirits-based canned cocktails, and hard seltzers belong to the brewery. Each will consent to the other's registration within its field. The negotiation technique, trading concessions that are cheap for the giver and valuable for the receiver, is the same as in any settlement, as described in the Settlement Negotiation Toolkit: Case Valuation Worksheets, Concession Plans, and Offer Letters.

Week 7: Drafting. The long-form agreement contains:

Recitals stating each party's marks, registrations, applications, years of use, channels, and the absence of known confusion since 2017.
A field-of-use allocation with definitions of "Wine Products" and "Brewery Products," and a procedure for new categories: a party wishing to use TIDEWELL on a product in neither field must give 60 days' notice, and the parties will consult in good faith; absent agreement, the product may be sold only under a different mark.
Presentation requirements: TIDEWELL BREWING or TIDEWELL with the lighthouse logo for the brewery; TIDEWELL CELLARS or TIDEWELL with the vineyard design for the winery; no use of the other's design elements or color schemes.
Channel commitments for the canned cocktails.
Mutual consents to use and register within each field, mutual covenants not to oppose, cancel, or otherwise challenge registrations within the other's field, and an obligation to sign any further consents the USPTO requires.
A cooperation clause: notify within 15 business days of any instance of actual confusion; meet within 30 days; take reasonable steps to remedy it.
An express statement that each party remains free to make and sell any product anywhere under marks other than TIDEWELL.
A no-license clause: neither party grants the other any right to use its marks; each uses its own mark as its own source identifier.
Successor and assignment terms, a notice-and-cure provision, a 30-day negotiation step and then mediation before suit, Maine law, and specific performance.

The short-form consent, intended for filing, repeats the recitals, the field line, the presentation and channel undertakings, and the cooperation clause. It does not say that confusion would be likely without the agreement. It says the parties believe confusion is unlikely because of the differences in their goods, presentation, and channels, and that they have agreed to the undertakings to preserve those differences.

Week 8: Fitting the register. Marisol amends Tidewell's identification to "prepared alcoholic cocktails, namely, spirits-based canned cocktails, excluding wine-based beverages." Daniel's client does not need to amend its registration. Marisol files the response with the short-form consent, a declaration from Tidewell's sales director about channels and customers, and photographs of both companies' packaging.

Month 3: The USPTO responds. The examining attorney withdraws the refusal, and the application is approved for publication. Had the refusal been maintained, Marisol's next step would have been a request for reconsideration with additional evidence and then an appeal, relying on Four Seasons and American Cruise Lines, with Mastic and Bay State distinguished.

Year 4: The acquisition. A national beverage company buys Tidewell Brewing. Its portfolio includes a wine-based canned spritzer sold under another brand, and its marketing team proposes relaunching the spritzer as TIDEWELL SPRITZ. The deal team's diligence found the coexistence agreement, and the acquirer's counsel flags it: wine-based beverages belong to the winery, the agreement binds successors, and a breach would expose the acquirer to an injunction and a contract claim. TIDEWELL SPRITZ is shelved. The spritzer stays under its existing brand. The agreement worked because it defined fields by what consumers buy (wine-based versus spirits-based beverages), not by the parties' then-current product lists, and because it bound successors.

Common mistakes

Filing a naked consent. A one-paragraph letter, or a list of DuPont factors with no facts, invites the Mastic result.
Conceding confusion. Agreements that describe a confusion problem the parties are managing, rather than differences that make confusion unlikely, hand the examining attorney the conclusion.
Using a consent to split territory. A promise to sell only in some states does not narrow a nationwide application. Use a concurrent use proceeding.
Leaving the identifications broad. The Board reads the identification as written. If the deal turns on a narrow product or channel, put it on the register where the rules allow.
Granting a license by accident. Permission to use "the other party's mark" is a license and needs quality control. Recognize each party's right to its own mark instead.
Dividing markets. Restrictions on where a party may sell or what it may sell under any name raise antitrust risk. Restrict use of the mark, not commerce.
Ignoring the future. New product categories, acquisitions, online sales, international expansion, and bankruptcy are where agreements fail. Provide for each.
Forgetting the agreement exists. Without a portfolio register and a launch review, the agreement will be breached by people who never heard of it.

Frequently asked questions

Will the USPTO accept any signed consent? No. A consent is evidence under the tenth DuPont factor. A detailed, bilateral agreement explaining why confusion is unlikely usually carries great weight; a bare letter may not. Examining attorneys and the Board still consider all of the other factors.

Do both parties have to sign? A registrant's letter of consent is enough to be considered, but a bilateral agreement signed by both parties is far more persuasive. Mastic treated a unilateral statement as a naked consent partly because it did not show any agreement between the parties.

Should the full agreement be filed with the USPTO? Not necessarily. Many parties file a short-form consent containing the recitals and undertakings that matter to registrability and keep commercial terms (payments, releases, dispute resolution) private. Anything filed becomes part of the public record.

Can a coexistence agreement be terminated? It depends on what it says. Many have no fixed term because the parties want permanent peace; others allow termination for uncured material breach, or on abandonment of a mark. Agreements that allow termination at will give little security to either side and usually carry less weight at the USPTO.

Is a coexistence agreement a license? Not if drafted correctly. Each party uses its own mark under its own control. If one party is allowed to use the other's mark, that part of the arrangement is a license and needs quality control.

Does the agreement protect us against third parties? No. It binds the parties and their successors. A third party's registration can still block an application, and third parties can still sue for infringement.

What happens if the other party goes bankrupt? The agreement is probably an executory contract that the debtor can assume or reject. After Mission Product Holdings, rejection is treated as a breach rather than a rescission, which should preserve rights already granted to the counterparty, but the application to coexistence covenants is not settled.

How long does the process take? A simple consent for an office action can be negotiated and filed within the response period, particularly with an extension. A full coexistence agreement resolving litigation may take one to three months. The step-by-step process is in Negotiating a Trademark Coexistence Agreement: A Practical Guide, and templates are in the Coexistence Agreement Toolkit: Term Sheets, Consent Agreements, and Field-of-Use Clauses.

Takeaways

The law favors private resolution of trademark conflicts. The USPTO, under DuPont and its progeny, gives great weight to thoughtful agreements between those who know the market best, and courts enforce them as contracts.
Weight depends on content. Facts, undertakings, cooperation, and a match between the agreement and the register make an agreement persuasive; conclusions and concessions make it weak.
The register is literal. Identifications and territorial scope must reflect the deal, and territorial splits need concurrent use registration.
Restrict the use of marks, not competition. Field and presentation terms tied to confusion survive antitrust scrutiny under the rule of reason; market division does not.
Keep the two sources distinct. Coexistence is not a license, and a license without control can destroy a mark.
Draft for the next owner, the next product, the next country, and the next crisis. That is where agreements succeed or fail.

This document is general information, not legal advice for any particular situation. Trademark, contract, antitrust, bankruptcy, and tax rules vary by jurisdiction and change over time; consult counsel about specific facts.