Stone Creek v. Omnia--Knowledge Destroys Good Faith Under the Tea Rose-Rectanus Doctrine

Picture this. You run a successful furniture company in Phoenix. You have spent two decades building a brand, advertising it, and earning customer loyalty around a distinctive red-oval logo. To make your higher-end pieces, you contract with an Italian leather manufacturer. The relationship hums along for years; they make beautiful sofas, you sell them, everyone is happy. Then one day a confused stranger in Wisconsin calls your Phoenix showroom asking about the warranty on a "Stone Creek" recliner you have never seen, sold at a department store you have never shipped to, made by a company you never authorized. You start pulling threads. The thread leads straight back to your own manufacturing partner—who, it turns out, has been quietly selling furniture under your brand, with your logo (recreated pixel by pixel from your own marketing materials, because a plain scan came out too blurry), to a Midwestern retail chain, for five years.

You sue. And here is the part that makes trademark lawyers sit up: your partner's lead defense is not "we didn't copy you." They plainly did. Their defense is a century-old doctrine most people have never heard of. It runs like this: "You never sold furniture in the Midwest. We built the customer base out there ourselves. Under the Tea Rose-Rectanus doctrine, we have a protected right to keep using that mark in that territory—and your later federal registration can't touch us."

That was the situation facing Stone Creek, Inc. when it discovered that Omnia Italian Design, Inc. had been selling leather furniture under the STONE CREEK brand to Bon-Ton department stores across seven Midwestern states. The resulting decision—Stone Creek, Inc. v. Omnia Italian Design, Inc., 875 F.3d 426 (9th Cir. 2017)—did something rare. It took a sleepy, hundred-year-old corner of trademark law, dragged it into the e-commerce era, and answered a question the Supreme Court had left dangling since 1918. The answer was blunt. Knowledge destroys good faith. If you knew about the other side's mark when you adopted it, you do not get the defense.

This article tells the whole story. We start with the two old Supreme Court cases that gave the doctrine its odd hyphenated name and explain, in plain English, what they held and why anyone designed a rule like this in the first place. We see how the Lanham Act of 1946 turned the map into "Swiss cheese," and how its twin engines—constructive notice and constructive use—decide who owns which holes. We walk through the facts of Stone Creek—a genuine business-partner betrayal—and the Ninth Circuit's three holdings, one of which the Supreme Court has since overturned. We map the real circuit split on what "good faith" means, with the cases on each side. And we close with worked hypotheticals, hard-nosed guidance for brand owners and litigators, and a short FAQ. By the end you will understand not just what Stone Creek held, but why a furniture dispute in Arizona quietly reshaped how territorial trademark rights are litigated nationwide.

The Two Cases Behind the Hyphen: Where the Doctrine Comes From

A world of local commerce

To understand Stone Creek, you have to understand the doctrine it interpreted, and to understand the doctrine you have to time-travel to a commercial world that barely exists anymore—a world without interstate highways, without national television advertising, certainly without Amazon, where a brand known in Boston might be utterly unknown in Birmingham, and where two honest merchants could each pour their life savings into building a business under the same name without either ever suspecting the other existed.

A trademark, at bottom, is a source identifier. It is the law's way of letting a buyer rely on a name or symbol to mean "this came from the maker I trust." Trademark law does not exist to reward whoever first scribbles a word on a sign; it exists to protect the goodwill—the earned reputation—that a name represents in the minds of actual customers, and to spare those customers the confusion of two different makers wearing the same name. (For the foundations, see our overview of the subject matter of trademark law, the substantive standards for protection, and our explainer on trademark rights under common law.) Because goodwill lives in customers' minds, and customers in 1900 were geographically scattered and largely unaware of distant brands, the early common law concluded—sensibly—that trademark rights were territorial. You owned the mark where your reputation actually reached, and not in the empty spaces beyond it.

That single insight is the seed of everything that follows. The "Tea Rose-Rectanus doctrine" is just the Supreme Court's two-part working-out of it.

Tea Rose: the flour case (1916)

The first half of the doctrine comes from Hanover Star Milling Co. v. Metcalf, 240 U.S. 403 (1916), universally nicknamed the "Tea Rose" case after the brand of flour at its center. One miller had been selling "Tea Rose" flour in Massachusetts, Ohio, and Pennsylvania since roughly 1872. Years later, a different company began selling "Tea Rose" flour deep in the South—Mississippi, Alabama, Georgia, Florida—with no knowledge whatsoever of the Northern miller. When the parties' commercial orbits finally brushed, the question was who owned "Tea Rose" in the South.

The Supreme Court refused to hand the Southern market to the first user simply because it had used the mark first somewhere. The right to a mark, the Court explained, grows out of its use, and a trademark "is not the subject of property except in connection with an existing business" (Hanover Star Milling, 240 U.S. at 413–14). A senior user who has never sold into a region, and whose reputation has never traveled there, cannot claim a monopoly over the name in that region against a good-faith newcomer who has built real goodwill there. The mark, in the South, meant the newcomer's flour—not the Northern miller's.

Rectanus: the drug case (1918), and the famous line

Two years later, United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90 (1918), supplied the doctrine's second half and its most-quoted language. One druggist had sold "Rex" remedies in New England since about 1877. Independently, and in genuine ignorance of the first, a Kentucky druggist named Theodore Rectanus had been selling his own "Rex" products in and around Louisville since roughly 1883, building a local following. When the New England company (by then United Drug) expanded into Kentucky and tried to shut Rectanus down on the strength of its earlier first use, the Court said no.

Justice Pitney's opinion is where the doctrine gets its philosophical spine. Trademark rights, he wrote, are not abstract property in a word floating free of commerce; they exist only to protect a business's actual goodwill. The senior user's prior use back East gave it no right to "monopolize markets that his trade has never reached, and where the mark signifies not his goods but those of another" (Rectanus, 248 U.S. at 98–100). To let a remote senior user reach into Louisville and seize the goodwill Rectanus had built with his own labor—goodwill that, to Louisville customers, was Rectanus—would not prevent confusion; it would manufacture injustice.

So the rule, distilled: a later user (a "junior user," in the trade) who adopts a mark (1) in good faith and (2) in a geographically remote market the senior user has not penetrated can acquire protectable rights in that remote market—rights even the senior user's earlier first use cannot override. Note carefully the two pillars. Remoteness is about geography. Good faith is about the junior user's state of mind. The Supreme Court was crystal clear about the geographic pillar. It was studiously vague about the second. In both Tea Rose and Rectanus, the junior users were genuinely innocent—they had never heard of the senior marks—so the Court never had to decide whether a junior user who did know could still claim "good faith." That unanswered question is the one Stone Creek would finally resolve a century later.

A clarifying point worth pausing on, because litigants confuse it constantly: Tea Rose-Rectanus is a defense, not a sword. It does not let a junior user expand or police a territory; it lets a junior user keep the limited turf it actually occupied, frozen at the boundaries it had reached. And it is a creature of common-law, unregistered rights, which arise from use in commerce rather than from any filing (see our note on proving trademark use and priority). The moment federal registration enters the picture, the analysis changes—which is exactly where the Lanham Act comes in.

The Lanham Act and the "Swiss Cheese" Map

The Lanham Act of 1946—still the core of federal trademark law—rebuilt the territorial landscape around two powerful ideas working in tandem: constructive notice and constructive use.

Start with constructive notice. Under 15 U.S.C. § 1072, registration of a mark on the Principal Register "shall be constructive notice of the registrant's claim of ownership." In plain terms, the day your mark hits the Principal Register, the entire country is charged by law with knowing you claim it—whether or not anyone actually checked. (We walk through how a mark gets there in the trademark process and the trademark registration guide; for keeping it alive afterward, see maintaining trademark registrations.)

Then add constructive use, the rule that pins down when nationwide rights attach. Under § 1057(c), 15 U.S.C. § 1057(c), filing an application on the Principal Register—even an intent-to-use application—gives the eventual registrant "a national right of priority" as of the filing date, against everyone except parties who already used the mark, or already filed, before that date. So the registrant's nationwide clock often starts ticking on the application date, not the registration date. For a junior user, the practical question is whether its honest, remote use predates the registrant's constructive-use (filing) date. If it does, the junior user may keep its island; if it does not, the registrant's nationwide priority swallows it.

Constructive notice and constructive use are the natural enemies of the Tea Rose-Rectanus defense, because the defense runs on innocence. A junior user who adopts a mark after the senior user has federally registered (or filed) cannot plausibly claim it was an unwitting, good-faith newcomer; the law deems it to have known. Federal registration, in effect, plants a flag over the whole map and says: no more honest, independent latecomers, anywhere.

But—and this is the wrinkle that gives the doctrine its modern life—the Lanham Act did not retroactively erase rights that already existed. Section 33(b)(5), 15 U.S.C. § 1115(b)(5), carves out a "limited area" exception. It preserves the rights of a junior user who, before the registrant filed (or before the registrant's constructive-use date), had already adopted the mark "without knowledge of the registrant's prior use" and continuously used it in a discrete territory. That junior user gets to keep its established area as a permanent island, immune even to the registrant's nationwide rights.

The vivid metaphor courts and commentators reach for is Swiss cheese. A federal registration gives you rights that blanket the whole country—but with holes punched out wherever some earlier good-faith junior user had already dug in before your registration. You own the cheese; they own the holes. And here is the statutory tell that Stone Creek would later seize on: even § 1115(b)(5) builds knowledge right into its text. The limited-area defense is available only to a junior user who acted "without knowledge of the registrant's prior use." The Lanham Act itself, in other words, treats knowledge as defense-defeating.

Freezing the map: what registration does to a senior user's own territory

There is a flip side worth understanding, because it shows how registration freezes everyone's rights, not just the latecomer's. Suppose the junior user is the one who registers first, while an unregistered senior user has been operating quietly in one region. Registration by the junior user does not extinguish the senior user's existing common-law rights (Dorpan, S.L. v. Hotel Meliá, Inc., 728 F.3d 55, 63 (1st Cir. 2013))—but it freezes the senior user's territory at whatever the senior user actually occupied as of the registration date (Allard Enterprises, Inc. v. Advanced Programming Resources, Inc., 249 F.3d 564, 572 (6th Cir. 2001); Tana v. Dantanna's, 611 F.3d 767, 780–81 (11th Cir. 2010)). The registrant takes the rest of the country; the senior user keeps only its frozen pocket (some courts allow it a narrow zone of natural expansion, e.g., Spiral Direct, Inc. v. Basic Sports Apparel, Inc., 293 F. Supp. 3d 1334, 1368 (M.D. Fla. 2017)). Each side can then sue the other for crossing the line. The lesson is the same from every angle: in the post-Lanham-Act world, the date someone files is the date the map gets drawn. Whoever registers first locks in the geography.

How the gap opened in Stone Creek

This is precisely the window Omnia tried to climb through. Stone Creek adopted its mark around 1990 and got an Arizona state registration in 1992, but did not federally register until 2012. Omnia's unauthorized Bon-Ton sales began in 2008—four years before that federal registration, and before any federal filing date. So Omnia could at least argue it slipped into a pre-registration gap where the common-law Tea Rose-Rectanus defense, not the stricter post-registration rule, governed. Whether it could actually invoke the defense came down to the one element the Supreme Court had never defined.

Everything turned on what "good faith" means.

The Facts: A Manufacturing Partner Goes Rogue

The facts of Stone Creek are as damning as they are instructive—a case study in how a trusted commercial relationship can curdle, and a standing argument for careful brand controls in any licensing or manufacturing deal. (The licensing dynamics here rhyme with those we cover in trademark overview: obtaining protection and licensing.)

Stone Creek was a furniture maker and retailer running five showrooms in the Phoenix area. Around 1990 it adopted the STONE CREEK mark—the words "Stone Creek" inside a distinctive red oval—for furniture. It secured Arizona state trademark protection in 1992 and a federal registration in 2012. Importantly, Stone Creek also advertised and sold furniture online, a fact that becomes pivotal later.

Omnia was an Italian leather furniture manufacturer. In 2003 the two companies struck a manufacturing deal: Omnia would produce furniture branded STONE CREEK for Stone Creek to sell. That arrangement handed Omnia an intimate education in Stone Creek's brand—its logo, its marketing materials, its product line, its customers. Omnia did not learn the STONE CREEK name from a trade show or a coincidence. It learned the name because it was on the inside.

In 2008—again, four years before Stone Creek's federal registration—Omnia began making STONE CREEK-branded furniture for Bon-Ton, a Midwestern department-store chain, with no knowledge or authorization from Stone Creek. And Omnia did not simply reuse a name. It reconstructed Stone Creek's logo. According to the record, because Omnia could not get a crisp image by scanning old documents, its people digitally recreated the identical red-oval mark from Stone Creek's own materials, then plastered it across binders, leather samples, color boards, and warranty cards—an entire turnkey brand presentation for Bon-Ton's furniture galleries in Illinois, Indiana, Iowa, Michigan, Ohio, Pennsylvania, and Wisconsin.

It ran for five years. Then, in 2013, came those warranty calls. Bon-Ton customers, holding cards printed with Stone Creek's real Phoenix address and phone number, dialed Stone Creek with questions about sofas Stone Creek had never built. The goodwill was so successfully transplanted that confused consumers found their way back to the genuine source—the clearest possible proof of confusion. Stone Creek sued in the District of Arizona for federal and common-law trademark infringement and unfair competition. (On how courts assess that kind of consumer-confusion proof, see navigating the maze of trademark confusion and our overview of infringement and related rights.)

The District Court Gets It Backwards

The trial court reached a result that, on these facts, reads almost upside down: no infringement. Its reasoning was geographic. Stone Creek's brick-and-mortar showrooms sat in Phoenix; Omnia's Bon-Ton sales happened in the Midwest; therefore, the court concluded, the parties operated in separate markets and consumers were not likely to be confused. Yet in the same breath the court found that Omnia could not invoke the Tea Rose-Rectanus defense, because Omnia plainly knew about Stone Creek's mark when it started.

Sit with the paradox. The court held there was no infringement because the markets were too remote for confusion—while simultaneously holding that the defense designed to protect remote good-faith users was unavailable because Omnia knew too much. Stone Creek had lost on the merits while winning the very point (knowledge defeats the defense) that should have spelled liability. Stone Creek appealed, and the Ninth Circuit took the opportunity to clean up both halves of the mess.

The Ninth Circuit: Three Holdings That Matter

Holding One: likelihood of confusion, reversed—and the internet changes the map

The Ninth Circuit reversed the no-confusion ruling, running the dispute through the Sleekcraft factors, its eight-factor framework for likelihood of confusion (AMF Inc. v. Sleekcraft Boats, 599 F.2d 341 (9th Cir. 1979)). The district court, the panel held, had suffered from a "myopic focus" on the physical distance between showrooms while ignoring almost everything else that mattered.

And almost everything else pointed one way. The marks were identical—not similar, identical, because Omnia had literally reconstructed the logo. The goods were identical: leather furniture versus leather furniture. The mark was conceptually strong: "Stone Creek" is arbitrary or fanciful as applied to furniture (it describes nothing about a sofa), placing it at the protected end of the distinctiveness spectrum that runs from generic to fanciful (a spectrum we map in our trademark basics primer). There was actual confusion, the gold standard of evidence—those warranty calls were confused consumers in the flesh. And Omnia's deliberate copying supplied the intent factor in the plaintiff's favor.

The most forward-looking move was what the panel did with geography. The district court had treated showroom location as decisive; the Ninth Circuit treated it as nearly beside the point, because Stone Creek sold online. In a world of e-commerce, the court reasoned, a Phoenix retailer's market is not bounded by the Phoenix metro area—it extends to anyone who can reach the website. Physical separation between two brick-and-mortar footprints no longer forecloses confusion when both sellers are a click away from the same shoppers. This was the doctrinal hinge of the case beyond the good-faith holding: it signaled that "geographic remoteness," the threshold Tea Rose-Rectanus requirement, is steadily eroding as commerce moves online. (For more on enforcing marks in digital channels, see brand protection online: a strategic guide for businesses.)

Holding Two: knowledge destroys good faith

Here is the heart of the opinion. To invoke Tea Rose-Rectanus, Omnia had to show good-faith use in a remote area. The good-faith element was the battleground, and the precise question was the one the Supreme Court had ducked in 1918: does a junior user's mere knowledge of the senior user's prior use, by itself, destroy good faith—or is knowledge just one ingredient in a broader inquiry into the junior user's intent?

This question had genuinely split the federal courts, and the Ninth Circuit had never weighed in. On one side, several courts treated knowledge as dispositive. The Seventh Circuit had put it plainly in Money Store v. Harriscorp Finance, Inc., 689 F.2d 666, 674 (7th Cir. 1982): "a good faith junior user is one who begins using a mark with no knowledge that someone else is already using it." The Eighth Circuit took the same view in National Association for Healthcare Communications, Inc. v. Central Arkansas Area Agency on Aging, Inc., 257 F.3d 732 (8th Cir. 2001), and the Trademark Trial and Appeal Board (TTAB) has long applied a knowledge-defeats-good-faith standard in concurrent-use proceedings. On the other side, courts treated knowledge as merely a factor, asking ultimately whether the junior user intended to trade on the senior user's goodwill. The Tenth Circuit took that path in GTE Corp. v. Williams, 904 F.2d 536 (10th Cir. 1990), and the Fifth Circuit has signaled the same in its Armand's Subway line of authority.

The Ninth Circuit came down hard on the majority side: knowledge of the senior user's prior use destroys good faith as a matter of law. Its reasoning braided two strands.

First, doctrinal history. The court returned to the source. In Tea Rose and Rectanus, the junior users were ignorant—that ignorance is precisely what justified protecting them. The good-faith requirement exists to shelter the merchant who independently builds a brand with no idea anyone else is using the name. That merchant deserves protection because, in its market, the goodwill it created really is its own. But the entire rationale collapses the instant the junior user knows. A knowing adopter is not independently building a brand; it is helping itself to one someone else already built. There is nothing to protect.

Second, and elegantly, the Lanham Act's own logic. Recall § 1115(b)(5): the statutory limited-area defense is open only to a junior user who acted "without knowledge of the registrant's prior use." And recall § 1072 and § 1057(c): federal registration and the underlying application impose constructive knowledge on everyone, nationwide, which is enough to slam the defense shut against later adopters. Put those together. If constructive knowledge—legal knowledge, fictional knowledge, knowledge you may not actually have—suffices to destroy the defense, then actual knowledge, the real article, must do at least as much. To hold otherwise would absurdly reward the junior user who genuinely knew over the one who merely should have known. The court refused. You cannot claim innocence, the logic runs, if you knew what you were doing.

Applied to Omnia, the conclusion wrote itself. Omnia was no stumbling innocent who chanced upon "Stone Creek" in a far-off market. Omnia was Stone Creek's manufacturing partner, had produced STONE CREEK furniture under contract for years, possessed Stone Creek's logo and materials, and had reconstructed that logo by hand. Its knowledge was not merely present; it was total. The Tea Rose-Rectanus defense was dead on arrival, and Omnia was liable for infringement.

Holding Three: willfulness for disgorgement—the holding the Supreme Court later abrogated

The third holding turned to remedies: did Stone Creek have to prove Omnia's infringement was willful to recover a disgorgement of Omnia's profits? The Ninth Circuit said yes, reaffirming its longstanding rule that willfulness is a prerequisite to a profits award under § 35(a) of the Lanham Act, 15 U.S.C. § 1117(a), and holding that Congress's 1999 amendments had not disturbed that rule.

That holding sat atop a second circuit split—a remedies split distinct from the good-faith one. When Congress amended § 1117(a) in 1999, it inserted the word "willful" expressly into the dilution pathway (tied to § 43(c), 15 U.S.C. § 1125(c)) while leaving the ordinary infringement pathway (§ 43(a), 15 U.S.C. § 1125(a)) textually untouched. Several circuits—including the Third, Fourth, Fifth, Sixth, Seventh, and Eleventh—read that surgical edit as a signal that willfulness was not required for profits in garden-variety infringement cases. Others—the First, Second, Eighth, Ninth, Tenth, and D.C.—kept the willfulness gate based on pre-1999 precedent and equity. Stone Creek planted the Ninth Circuit firmly in the willfulness-required camp and remanded for a determination of whether Omnia's conduct cleared that bar.

The update that rewrites this holding. Three years later, the Supreme Court resolved the remedies split in Romag Fasteners, Inc. v. Fossil, Inc., 590 U.S. 212, 140 S. Ct. 1492 (2020). Writing for a unanimous Court, Justice Gorsuch held that willfulness is not an inflexible precondition to a profits award under § 1117(a). The text, he observed, "has never required a showing of willfulness to win a defendant's profits," and the Court declined to read into the statute "words that aren't there"—particularly given that the Lanham Act uses the word "willful" expressly elsewhere, proving Congress knew how to impose such a requirement when it wanted to (Romag, 140 S. Ct. at 1495–97). But Romag did not make mental state irrelevant. The Court stressed that "a trademark defendant's mental state is a highly important consideration in determining whether an award of profits is appropriate." Willfulness moved from a threshold gate to a heavy thumb on the equitable scale (see, summarizing the post-Romag landscape, Trademark Infringement and Dilution Claims, Remedies, and Defenses, Practical Law Practice Note (collecting circuit standards on willfulness and profits)).

So Stone Creek's third holding is, today, no longer good law on this point. Disgorgement is now available in every circuit without a categorical willfulness showing—though a defendant's culpability still powerfully shapes whether, and how much, a court will order. (Willfulness remains a strict precondition only on the dilution side under § 1125(c).) For practitioners, Romag meaningfully raised the value of disgorgement as a remedy. In many trademark cases the plaintiff's own lost-profits or actual damages are maddeningly hard to quantify—how much revenue did you lose specifically because of the infringement, as opposed to ordinary market noise? The defendant's profits are often the only number a court can fix with confidence, and dropping the willfulness gate put that recovery within reach even where the infringer acted with "callous disregard" rather than provable intent.

A Worked Hypothetical: Meet Acme Coffee and Brewster's

Doctrine sticks better with an example, so let us invent two clean ones. (Both are entirely hypothetical and invented for illustration.)

Hypothetical 1 — the genuine innocent. Acme Coffee Roasters opens in Portland, Oregon in 2010, selling "ACME" coffee from a single café, never advertising beyond Oregon and never online. In 2014, in Atlanta, an unrelated entrepreneur named Dana opens "ACME" coffee shops, having never heard of the Portland Acme—Dana picked the name as a wink at the old cartoons. Neither sells into the other's region; neither knows the other exists. In 2018, Portland Acme federally registers and, expanding east, demands Dana stop. Classic Tea Rose-Rectanus. Dana adopted in genuine good faith, in a remote market, before Portland's 2018 filing. Under § 1115(b)(5) and the common-law doctrine, Dana likely keeps Atlanta as a frozen "Swiss-cheese hole"—she cannot expand under the name, but she need not abandon the goodwill she honestly built. This is the doctrine doing exactly what it was designed to do.

Hypothetical 2 — the Stone Creek problem. Same Portland Acme. But now suppose Dana had spent two years as Portland Acme's wholesale distributor, handling its bags, its logo, its customer lists—and then quietly opened her own "ACME" shops in Atlanta using a logo she rebuilt from Portland's materials. Now Dana knew. In the Ninth, Seventh, or Eighth Circuit, her Tea Rose-Rectanus defense is gone before it starts: knowledge is dispositive, full stop. In the Fifth or Tenth Circuit, Dana could at least argue good faith—but on these facts (a former insider rebuilding the senior user's exact logo) even an intent-focused court would almost certainly find she meant to trade on Portland's goodwill. The circuit split changes the route to the result; on egregious facts like Stone Creek's, it rarely changes the destination.

Where the split actually bites is the closer middle case—the junior user who knew of the mark casually but adopted it for an independent, arguably innocent reason (say, a common geographic or descriptive term) and never tried to free-ride. Hypothetical 3 makes that concrete: imagine "Cedar Ridge" hardware stores in rural Vermont and, years later and 1,200 miles away, an unrelated "Cedar Ridge" hardware store in a Cedar Ridge subdivision of suburban Texas, whose owner had once glimpsed the Vermont chain's website but chose the name simply because that is the neighborhood's name. In a knowledge-is-dispositive circuit, the Texas store's defense dies on the glimpse alone. In an intent-focused circuit, it may well survive, because the owner can show an honest, independent reason for the name and no design on the senior user's goodwill. Same facts, opposite outcomes—decided entirely by where the suit is filed.

That middle case is why forum matters, and why this otherwise dusty doctrine still draws blood.

The Good-Faith Circuit Split, Mapped

The split that Stone Creek sharpened is alive and unresolved. The Supreme Court denied certiorari in Stone Creek in 2018, declining once again to settle the meaning of "good faith" it left open in 1918.

Circuit / Body	Standard	Key Authority
Seventh Circuit	Knowledge destroys good faith	Money Store v. Harriscorp Finance, 689 F.2d 666 (7th Cir. 1982)
Eighth Circuit	Knowledge destroys good faith	Nat'l Ass'n for Healthcare Commc'ns v. Cent. Ark. Area Agency on Aging, 257 F.3d 732 (8th Cir. 2001)
Ninth Circuit	Knowledge destroys good faith	Stone Creek v. Omnia, 875 F.3d 426 (9th Cir. 2017)
TTAB	Knowledge destroys good faith	Consistent agency position in concurrent-use practice
Fifth Circuit	Knowledge is one factor; ask whether the junior user intended to exploit the senior's goodwill	Armand's Subway line of authority
Tenth Circuit	Knowledge is one factor; intent-focused	GTE Corp. v. Williams, 904 F.2d 536 (10th Cir. 1990)

In the Seventh, Eighth, and Ninth Circuits (and at the TTAB), the rule is a bright line: a junior user who knew of the senior user's mark cannot claim Tea Rose-Rectanus, period. Knowledge ends the inquiry. In the Fifth and Tenth Circuits, knowledge is evidence but not a verdict; a knowing junior user can still argue good faith by showing it did not set out to siphon the senior user's goodwill—perhaps because it adopted a descriptive or geographic term for honest reasons, or used the mark on different enough goods in a different enough market that confusion never crossed its mind.

The practical upshot is forum sensitivity. Where a dispute could be filed in more than one circuit, the choice of forum can decide the case. And the academic verdict has largely favored Stone Creek: a Boston College Law Review analysis concluded the Ninth Circuit got it right, because a knowledge-based standard "reinforces trademark law's ultimate policy goals of preventing unfair competition and consumer confusion" (Roya Tabibi, Good Faith Trademark Infringement, 60 B.C. L. Rev. E. Supp. II.-65 (2019)). That is the better reading: the majority rule both honors the doctrine's original innocence-protecting purpose and harmonizes the common-law standard with the Lanham Act's own knowledge-keyed text.

A note for the historically minded: this good-faith question is a close cousin of trademark law's other great territorial puzzle, the Dawn Donut rule (Dawn Donut Co. v. Hart's Food Stores, Inc., 267 F.2d 358 (2d Cir. 1959)), under which a federal registrant operating only in a distant region may be denied an injunction against a remote junior user until the registrant actually moves into the junior user's territory—because, until then, there is no likely confusion to enjoin. Dawn Donut and Tea Rose-Rectanus are flip sides of the same coin: one constrains the registrant's sword in remote areas, the other shields the unregistered junior user's shield. And both, tellingly, are under pressure from the same force—the internet—that Stone Creek invoked. Several judges have openly questioned whether Dawn Donut survives a world where every registrant is, in some sense, "present" everywhere online.

Why Stone Creek Matters More Than Ever

The internet is quietly euthanizing "remoteness"

The Tea Rose-Rectanus doctrine was engineered for a world in which a flour brand in Massachusetts and a flour brand in Mississippi could coexist in total mutual ignorance. That world is gone. E-commerce, search advertising, social media, and online marketplaces have collapsed the very distances the doctrine presupposes. A Phoenix furniture company sells to Wisconsin through its website; a single-café roaster in Portland builds a national following on Instagram; a regional brand becomes searchable, shoppable, and confusable from every zip code at once.

Stone Creek baked that reality into its confusion analysis. Because Stone Creek sold online, the panel held, the physical gap between Arizona showrooms and Midwestern department stores did not prevent confusion. The implication for the doctrine is profound and runs independently of the good-faith holding: if online commerce makes almost no market truly "remote," then the doctrine's threshold geographic requirement becomes harder and harder to satisfy—no matter how pure the junior user's intentions. A junior user might be a saint, ignorant of the senior mark and innocent of any free-riding, and still lose because, in 2026, it cannot honestly say it operated in a market the online senior user had not reached. The good-faith debate is, in a sense, becoming a debate about a shrinking island.

For brand owners, the lesson is liberating: geographic separation protects you far less than it once did, but it also shields infringers far less. (For the digital-enforcement toolkit, see brand protection online and our guide on filing a UDRP complaint for domain-name disputes.)

The licensee/manufacturer trap

The most unnerving feature of Stone Creek is that the villain was not a stranger but a partner. Manufacturing agreements, OEM deals, private-label arrangements, and distribution contracts all hand a third party the keys to your brand: your logo files, your marketing assets, your customer relationships, your reputation. Most partners are honest. But the structure of the relationship is exactly what made Omnia's betrayal possible—and so easy. Omnia did not have to invent a brand; it had been handed one.

The defense against the Omnia scenario is contractual, and it must be built before the relationship sours, not after. A well-drafted trademark license or manufacturing agreement should, at minimum, define the scope of authorized use with precision; restrict the licensee to specified products, channels, and territories; require prior written approval for every use of the mark; impose strict quality-control obligations (the linchpin of any valid license—a license without genuine quality control can be voided as a "naked license," a trap we discuss in trademark overview: obtaining protection and licensing); reserve audit and inspection rights; grant immediate termination on unauthorized use; and—critically—impose post-termination duties to cease use and to destroy or return all branded materials, certified in writing. These are not boilerplate niceties. They are the evidentiary and contractual scaffolding that makes enforcement swift and decisive when a partner goes rogue. Without them, you are left holding only trademark law's after-the-fact remedies, deployed against damage already done. (For drafting confidentiality and IP protections into commercial deals more broadly, see drafting enforceable non-disclosure agreements for technology transactions.)

Register early, and register federally

Stone Creek is also a quiet sermon on the cost of registration procrastination. Stone Creek used its mark from around 1990 and held Arizona state rights from 1992—but waited until 2012 to register federally. That twenty-year gap is the whole ballgame. Because Omnia's unauthorized use began in 2008, before the 2012 federal registration and any federal filing, Omnia could at least mount the Tea Rose-Rectanus argument and force years of litigation over "good faith." Had Stone Creek registered federally in, say, 1995, the constructive notice of § 1072 (anchored to the constructive-use date under § 1057(c)) would have charged Omnia—and everyone else—with knowledge as of that date, and the limited-area defense would have been categorically off the table. No registration, no constructive notice, no automatic disqualification—exactly the gap Omnia tried to exploit.

The economics are lopsided. A federal application costs a few hundred dollars per class plus modest attorney time—a rounding error against a single day of trademark litigation. In exchange you get nationwide constructive notice, a constructive-use priority date as of filing, the presumptions of validity and ownership, the path to incontestability after five years, access to federal court, statutory remedies, and the power to record the registration with U.S. Customs and Border Protection to block counterfeit imports. Above all, federal registration slams the Tea Rose-Rectanus door on every future user. State registration—useful as it is—does none of this on a national scale. (Start with how to file a trademark application with the USPTO and the federal trademark application checklists; clear the path first with how to conduct a comprehensive trademark clearance search, and time the filing using our note on when to trademark your brand.)

The Good-Faith Theme Beyond Territory: A Note on Willfulness and the Clearance "Shield"

Stone Creek is, in a sense, two stories about good faith wearing one caption. The first is the territorial story we have told: knowledge defeats the Tea Rose-Rectanus defense. The second is the remedies story: the defendant's mental state shapes the profits remedy. Romag untethered those two—knowledge can defeat the defense in one breath while not automatically being required for disgorgement in the next—but in practice they often move together, because the same evidence that proves a junior user "knew" (and thus loses the defense) tends also to prove the kind of culpability that pushes a court toward awarding profits.

This is the mirror image of a defense-side strategy worth flagging: the good-faith shield that a thorough clearance search, a competent attorney opinion, and USPTO approval can provide against a charge of willful infringement. A defendant who searched diligently, relied on reasoned advice, and adopted in honest reliance on a clean clearance is far better positioned to argue it acted in good faith—both to defeat enhanced or punitive exposure and to influence the court's Romag equitable calculus. We develop that defensive playbook in the shield of good faith: how trademark clearance searches, attorney opinions, and USPTO approval can protect against willful infringement claims. The pairing is instructive: Stone Creek shows how knowledge can be a junior user's undoing, while the clearance shield shows how the absence of bad knowledge—documented and reasoned—can be a junior user's salvation. Good faith, in trademark law, is a fact you build or destroy with your own conduct and records.

Practical Guidance

For brand owners

Combine three habits and the Omnia scenario becomes very hard to inflict on you. Register early and federally, ideally before you enter any licensing, manufacturing, or distribution relationship that hands outsiders your brand. Monitor relentlessly—USPTO filings (watch services help), online marketplaces, social platforms, app stores, and domain registrations—because in a borderless market infringement can erupt anywhere. And paper your relationships with agreements that define authorized use narrowly, restrict territory and channels, and arm you with audit, approval, and termination rights.

When you spot unauthorized use, move fast. Delay is doubly costly: it lets the junior user accumulate the very independent goodwill that strengthens its territorial argument and softens a court's appetite to enjoin, and it can fuel laches or acquiescence defenses. (We cover those equitable defenses in understanding equitable defenses: laches, acquiescence, waiver, and equitable estoppel.) A prompt, well-aimed cease-and-desist letter—see drafting a trademark cease and desist letter—often resolves matters before litigation, and even where it does not, it builds the record.

For accused junior users

If you are charged with infringement and believe you hold honest territorial rights, the circuit you are in may decide your fate. In the Seventh, Eighth, or Ninth Circuit, any credible knowledge of the senior user's prior use—however casual—sinks your Tea Rose-Rectanus defense. In the Fifth or Tenth Circuit, knowledge is survivable if you can show you never intended to exploit the senior user's goodwill. In every circuit, the defense requires that your use predate the senior user's federal registration—and, more precisely, its constructive-use (filing) date under § 1057(c); if the federal filing came first, constructive notice charges you with knowledge as a matter of law and the limited-area defense is unavailable regardless of your actual innocence. Marshal your evidence of independent adoption early: dated invoices, advertising, internal memos showing why you chose the name, and proof of the territory you actually occupied.

For litigators

In knowledge-is-dispositive circuits, knowledge is your spear or your shield, and discovery should be built around it. Manufacturing and distribution agreements, emails, trade-show records, industry-publication subscriptions, web-search histories, and prior dealings between the parties can all establish—or negate—knowledge at the moment of adoption. Because knowledge is a factual question that documents often answer cleanly, the Tea Rose-Rectanus defense can frequently be won or lost on summary judgment, sparing the expense of trial.

Do not neglect the territorial-proof half of the case, either. Whether a market is "remote" and how far each party's rights actually reach is a fact question proved with use evidence—sales records, advertising expenditures and reach, geographic distribution, and unsolicited publicity—often packaged in a detailed use declaration. Where the senior user's rights were frozen by an intervening registration (Allard, Tana), pin down the territory as of that date. Where authenticating the digital paper trail matters (and it usually does), see our practitioner's guide to authenticating website screenshots as evidence in federal court.

On remedies, remember the Romag recalibration: you no longer must prove willfulness to open the door to disgorgement, but the defendant's mental state is "a highly important consideration," so develop the willfulness record anyway. A court confronted with evidence of deliberate copying—an Omnia-style hand-reconstructed logo—is dramatically more likely to exercise its equitable discretion in favor of a profits award, and to consider enhancement and an exceptional-case fee award under § 1117(a). Build the culpability narrative even though, post-Romag, it is no longer a strict gate.

Key Takeaways

The doctrine in one line. A junior user who adopts a mark in good faith and in a remote market the senior user has not reached can acquire protectable common-law rights there—a Tea Rose-Rectanus "hole" in the senior user's "Swiss cheese" map.
The Stone Creek rule. In the Ninth Circuit, a junior user's knowledge of the senior user's prior use destroys "good faith" as a matter of law—aligning with the Seventh and Eighth Circuits and the TTAB, and against the more lenient, intent-focused Fifth and Tenth Circuits.
Statutory backbone. The holding tracks the Lanham Act: § 1072 imposes nationwide constructive notice on registration, § 1057(c) sets the constructive-use priority date at filing, and § 1115(b)(5) limits the defense to junior users who acted "without knowledge." If constructive knowledge defeats the defense, actual knowledge must too.
The internet erodes "remoteness." Because online selling reaches everywhere, the doctrine's geographic threshold is increasingly hard to meet—independent of the good-faith question.
Remedies update. Stone Creek required willfulness for disgorgement, but Romag Fasteners v. Fossil (2020) abrogated that requirement for infringement; mental state remains a heavyweight equitable factor, not a gate (willfulness is still required for dilution profits).
Prevention beats litigation. Register federally early; license and manufacture under tight, audited, terminable agreements; monitor; and enforce promptly.

Frequently Asked Questions

What is the Tea Rose-Rectanus doctrine, in plain English? It is a rule that common-law trademark rights are territorial. A later user who, in good faith and without knowing about an earlier user's mark, builds a real business under the same name in a different region the earlier user never reached can keep using the mark in that region—even against the earlier user. It is named for two Supreme Court cases, the 1916 "Tea Rose" flour case (Hanover Star Milling v. Metcalf) and the 1918 "Rex" drug case (United Drug v. Rectanus).

What exactly did Stone Creek v. Omnia decide? Three things. (1) It reversed a no-confusion ruling, holding that identical marks on identical goods, plus actual confusion and online sales, create likely confusion despite physical distance between stores. (2) It held that a junior user's knowledge of the senior user's mark destroys the good faith Tea Rose-Rectanus requires—so Omnia, Stone Creek's own manufacturing partner, could not use the defense. (3) It required willfulness for a profits award—a holding the Supreme Court later abrogated in Romag.

Is Stone Creek still good law? The good-faith holding (knowledge destroys good faith) is still good law in the Ninth Circuit and remains influential. The willfulness-for-disgorgement holding is not good law after Romag Fasteners, Inc. v. Fossil, Inc., 590 U.S. 212, 140 S. Ct. 1492 (2020), which held that willfulness is not a precondition to a profits award in an infringement case.

Does federal registration eliminate the Tea Rose-Rectanus defense? Effectively, yes—going forward. Federal registration gives nationwide constructive notice under 15 U.S.C. § 1072 and a constructive-use priority date as of filing under § 1057(c), so anyone who adopts the mark after that date is charged with knowledge and cannot claim the innocent, good-faith status the defense requires. The defense survives only for junior users who were already using the mark, in good faith, in a defined area before that date (15 U.S.C. § 1115(b)(5)). That is the strongest practical argument for registering early.

Which circuits say knowledge destroys good faith, and which do not? Knowledge is dispositive in the Seventh (Money Store), Eighth (Healthcare Communications), and Ninth (Stone Creek) Circuits, and at the TTAB. Knowledge is treated as just one factor—focused on intent to exploit goodwill—in the Tenth (GTE v. Williams) and Fifth (Armand's Subway line) Circuits. The Supreme Court has not resolved the split and denied certiorari in Stone Creek in 2018.

My contract manufacturer or licensee is selling under my brand without permission. What can I do? Act immediately. You likely have federal and common-law infringement claims; if the use began after your federal filing, the Tea Rose-Rectanus defense is unavailable to them. Preserve evidence of their knowledge (your agreement alone usually proves it), send a cease-and-desist, and consider a TRO or preliminary injunction to halt ongoing sales. Going forward, ensure your agreements include territorial and channel restrictions, prior-approval and audit rights, quality control, and post-termination destruction-of-materials obligations.

Does the doctrine even matter now that everyone sells online? Less and less, and that is part of Stone Creek's point. The defense requires a "remote" market, and online commerce makes nearly every market reachable, so the geographic threshold is increasingly hard to satisfy. The doctrine still matters for purely local, pre-registration, brick-and-mortar uses—and for the rich body of priority and concurrent-use principles it anchors—but its real-world footprint is shrinking.

What is the difference between Tea Rose-Rectanus and the Dawn Donut rule? They are flip sides of trademark territoriality. Tea Rose-Rectanus shields an unregistered junior user's established turf against a senior user. The Dawn Donut rule (2d Cir. 1959) limits a federal registrant's ability to win an injunction against a remote junior user until the registrant actually enters that territory, because absent overlap there is no likely confusion to enjoin. Both rules are under pressure from the same internet-driven erosion of "remoteness."

If a junior user registers first, does the original senior user lose everything? No. The senior user keeps its existing common-law rights, but a junior user's registration freezes the senior user's territory at whatever it actually occupied as of the registration date (Allard Enterprises v. Advanced Programming Resources; Tana v. Dantanna's). The registrant takes the rest of the country, and each side can sue the other for crossing the line. Some courts allow the senior user a narrow zone of natural expansion, but the safe assumption is that the map freezes when someone files.

This article is for general informational purposes only and does not constitute legal advice. Trademark rights, defenses, and remedies depend on the specific facts and the law of the applicable jurisdiction, and the circuit split discussed here means outcomes can differ by forum. Consult qualified trademark counsel regarding your specific circumstances.