Few obligations in patent law are as easy to satisfy, as cheap to implement, and as financially devastating to neglect as patent marking. A company can spend half a million dollars prosecuting a patent, build a business around the covered product, sell that product for years without ever stamping a patent number on it, and then walk into court to discover that the law forbids recovery of a single dollar of damages for all of those years of infringement — even where the infringer copied the product on purpose. The cure cost a few cents per unit. The omission can cost millions.

That asymmetry has only grown sharper. For more than a century, an infringer who had slept on a patentee's silence could sometimes lean on the equitable defense of laches to trim a stale damages claim, and a patentee who forgot to mark might at least argue the equities cut both ways. No longer. In SCA Hygiene Products Aktiebolag v. First Quality Baby Products, LLC, 580 U.S. 328 (2017), the Supreme Court abolished laches as a defense to patent damages claims brought within the six-year window of 35 U.S.C. § 286. With laches gone, the marking statute, 35 U.S.C. § 287, is now the principal — often the only — doctrine that determines how far back a patentee can reach for money. Marking went from important to indispensable.

This article is a focused, in-depth treatment of patent marking under § 287: the section of the Patent Act that governs how a patentee gives the public notice of its rights and when damages begin to run. It is the deep, patent-only counterpart to our broader overview, Understanding Patent and Trade Dress Marking Requirements -- A Comprehensive Guide for Businesses and Practitioners, which treats patent marking and trademark/trade dress marking together. That companion piece is the right place to start if you also sell branded goods and need the whole notice landscape in one view; the two regimes look superficially similar but run on opposite logic, a contrast we draw out below. This article assumes you came for the patent half and wants to go deeper: the constructive-notice mechanism, the actual-notice alternative, the America Invents Act's virtual-marking option, the patentee's duty to police its licensees, the special problems posed by method claims and multi-patent products, the diminished but still-real risk of false marking under 35 U.S.C. § 292, and a concrete compliance program you can actually run.

The discussion is written so that a judge, a seasoned patent litigator, an in-house generalist, and an entrepreneur with a single issued patent can all follow it. Terms of art are explained in plain language the first time they appear, and the law is illustrated with worked hypotheticals using clearly labeled invented parties.

What Patent Marking Is, and Why It Exists

To "mark" a patented article means to put the public on notice that the article is covered by one or more patents, typically by affixing the word "patent" (or the abbreviation "pat.") together with the patent number to the product itself. Marking is not a condition of patent validity. A patent is fully valid and enforceable whether or not the patented product is marked, and an unmarked patentee can still sue and win on liability. What marking does is govern the timing of damages — it determines how far back in time a patentee can reach to recover money for infringement. Liability and damages are separate questions, and marking lives entirely in the second one.

The Supreme Court has explained that Congress designed the marking system to provide "a ready means of discerning the status of intellectual property embodied in an article of manufacture or design." Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 162 (1989). The statute serves three intertwined purposes the Federal Circuit has repeatedly identified: (1) it helps avoid innocent infringement by alerting the public that an article is patented; (2) it encourages patentees to give public notice that the article is patented; and (3) it aids the public in identifying whether a given article is patented. Nike, Inc. v. Wal-Mart Stores, Inc., 138 F.3d 1437, 1443 (Fed. Cir. 1998). In short, marking channels information from the patentee to the marketplace, so that competitors, customers, and would-be designers-around can make informed decisions before they commit capital.

Because the doctrine is fundamentally about notice, keep one question in front of you at all times: as of any given date, did the accused infringer have notice that the article was patented? The Patent Act recognizes two ways the answer can be "yes" — constructive notice and actual notice — and the difference between them is the heart of § 287.

It is worth pausing on how this contrasts with trademark and trade dress, the subject of our companion guide. The federal trademark registration symbol (®) and the © copyright notice operate on a permissive, reward-the-diligent logic: using them unlocks remedies (such as certain damages or a presumption of notice) but failing to use them rarely zeroes out recovery outright. Patent marking is harsher. Under § 287 the default for an unmarked patentee is zero pre-notice damages, and the patentee must affirmatively earn its way back to recovery. The trademark owner who forgets the ® loses an enhancement; the patent owner who forgets the number can lose the case's entire pre-suit damages. That difference in default rules is the single most important reason patent marking deserves its own treatment.

The Statutory Text and the Two Forms of Notice

Section 287(a) provides, in relevant part, that patentees and persons making, offering for sale, or selling a patented article within the United States, or importing it, "may give notice to the public that the same is patented" either by marking the article with the word "patent" or "pat." (or the virtual-marking URL discussed below) together with the patent number, or, when the article cannot be marked, by marking the package. The statute then delivers the critical consequence: "In the event of failure so to mark, no damages shall be recovered by the patentee in any action for infringement, except on proof that the infringer was notified of the infringement and continued to infringe thereafter, in which event damages may be recovered only for infringement occurring after such notice."

That single sentence builds the entire framework. Read it slowly. If the patentee fails to mark, no damages are recoverable — full stop — except on proof that the infringer was notified of the infringement and continued anyway, in which case damages run only from the date of that notification forward. Marking is therefore the patentee's mechanism for collecting damages from the very first infringing sale; actual notice is the fallback when the patentee did not mark. The two are not additive — they are alternative routes to the same destination, and the patentee needs only one of them to keep its damages alive.

Constructive Notice: Marking the Product

When a patentee properly and consistently marks its patented articles, the law treats the entire world as being on notice of the patent. This is constructive notice — "constructive" because no one needs to have actually seen the marking; the act of marking is deemed, as a matter of law, to give notice to everyone. The payoff is enormous: a patentee that has marked may recover damages for infringement occurring at any time within the statutory damages window, which under § 286 reaches back up to six years before the complaint is filed, regardless of when (or whether) the infringer actually learned of the patent.

Constructive notice is the prize. It decouples damages from the infringer's state of mind. The infringer cannot defeat pre-suit damages by claiming ignorance, because the patentee did its part by marking. And after SCA Hygiene, that six-year reach is not subject to equitable shortening for delay — laches cannot bar damages within § 286's window — so a marking patentee who waits a few years to sue does not forfeit the back damages it would otherwise be owed. SCA Hygiene, 580 U.S. at 346-48. Marking thus does double duty: it captures pre-suit damages and it makes the full statutory window genuinely available.

There is one important category of patentee for whom the marking duty does not arise at all: the patentee who neither makes the patented article itself nor authorizes anyone else to make it. The Supreme Court held long ago that a patentee that has neither manufactured nor licensed the manufacture of any patented article may recover full damages without marking, because the marking obligation is keyed to the existence of a tangible article that can carry the mark. Wine Railway Appliance Co. v. Enterprise Railway Equipment Co., 297 U.S. 387, 395-401 (1936). The Court reasoned that the statute was not meant "to impose a new and different burden upon non-producing patentees" or to deprive them of their common-law right to full recovery; where no article is made or sold, there is nothing to mark, and the actual-notice fallback becomes relevant only when the first form of notice — the visible mark — could have been given but was not. A pure licensing or non-practicing patentee that sells nothing (and has no licensee selling anything) therefore stands much like a method-only patentee: with no marked-or-unmarked article in the world, constructive notice is, in effect, automatic. The risk reappears the moment the patentee or any licensee actually sells a covered product, because then a markable article exists and § 287 attaches to it.

Actual Notice: The Specific Charge of Infringement

If the patentee did not mark — or did not mark consistently enough to earn constructive notice — its only route to pre-suit damages is actual notice under the statutory exception. Here practitioners must be precise, because "actual notice" in § 287 is a term of art that means far more than the infringer happening to know the patent exists.

The Federal Circuit has held that actual notice requires "the affirmative communication of a specific charge of infringement by a specific accused product or device." Amsted Industries Inc. v. Buckeye Steel Castings Co., 24 F.3d 178, 187 (Fed. Cir. 1994). Three points follow. First, notice must come from the patentee (or someone acting for it), not from something the infringer learned on its own. The focus is on the patentee's actions, not the infringer's knowledge. Second, it is not enough to tell a competitor that a patent exists or even to offer it a license; the patentee must identify the patent and accuse a specific product of infringing it. Third, the infringer's subjective awareness is irrelevant — an infringer who knew about the patent for years owes nothing for that period unless the patentee affirmatively charged it with infringement.

This is counterintuitive and trips up clients constantly. A patentee may have sent the infringer a friendly note saying "we hold U.S. Patent No. X, would you like to license it?" and feel certain the infringer was "on notice." But that letter, lacking a specific accusation against a specific product, does not start the damages clock. The Federal Circuit reaffirmed the date-specific, charge-specific nature of actual notice in Funai Electric Co. v. Daewoo Electronics Corp., 616 F.3d 1357, 1373 (Fed. Cir. 2010), where it confirmed that for unmarked products, damages are measured from the date the accused infringer received notice identifying the patent and the accused product — not from some earlier, vaguer contact. The classic safe formulation is a letter that names the patent, names the accused product, and states that the product infringes. (For the mechanics of such letters, including the careful drafting needed to avoid triggering a declaratory-judgment action, see the discussion below and our Comprehensive Guide to Patent Infringement Litigation -- From Summary Judgment Denial to Post-Trial.)

A related trap concerns who must give the notice. Because § 287 focuses on the patentee's conduct, notice from a stranger — an industry newsletter that reports the patent, or the infringer's own clearance counsel who flags it during a freedom-to-operate review — does not satisfy the statute. The communication must emanate from the patentee or someone authorized to act for it. Nor does it matter that the infringer plainly knew it was copying; the Federal Circuit has been emphatic that "it is irrelevant ... whether the defendant knew of the patent or knew of his own infringement. The correct approach to determining notice under section 287 must focus on the action of the patentee, not the knowledge or understanding of the infringer." Amsted, 24 F.3d at 187. This is why a deliberate copyist can owe nothing for years: the law conditions pre-notice damages for unmarked products on the patentee speaking up, and silence has consequences no matter how culpable the infringer. (On what copying actually amounts to as a matter of liability, see What Constitutes Patent Infringement.)

The content of the notice also matters. A bare reference to a patent number, without an accusation, is insufficient; so is a vague assertion that "your products may infringe our portfolio." The communication should identify the specific patent (by number) and the specific accused product or product line, and convey that the patentee believes the identified product infringes. The notice need not threaten suit, demand a particular royalty, or rise to an ultimatum — it must simply communicate a charge of infringement. But the line between an effective charge and a benign licensing overture is thin, and crossing it can hand the recipient a basis to file a declaratory-judgment action in a forum of its choosing, which is exactly the subject of the next section.

Because actual notice is so demanding and so date-specific, the single most reliable way to maximize damages is to mark from the beginning so that constructive notice is in place and the actual-notice inquiry never matters. Filing suit is itself a form of actual notice — service of a complaint that pleads infringement satisfies the requirement — but that captures damages only from the filing date forward, leaving every earlier infringing sale unrecoverable if the product was not marked.

The Notice-Letter Dilemma: Damages Clock Versus Declaratory-Judgment Risk

There is a built-in tension in the actual-notice route that deserves its own discussion, because it is where good intentions go wrong. A patentee that did not mark has every incentive to send a sharp notice letter, because the sharper and more specific the charge, the more clearly it starts the damages clock. But the same specificity that satisfies § 287 can create declaratory-judgment jurisdiction, letting the accused infringer race to its preferred court and sue first for a declaration of non-infringement, invalidity, or unenforceability.

The governing standard comes from MedImmune, Inc. v. Genentech, Inc., 549 U.S. 118, 127 (2007), which loosened the test for declaratory-judgment standing: a justiciable controversy exists when, under all the circumstances, there is a substantial dispute of sufficient immediacy and reality — the recipient no longer needs to wait under threat of imminent suit. The Federal Circuit has applied this to hold that a patentee's conduct showing an intent to enforce can support jurisdiction. See Asia Vital Components Co. v. Asetek Danmark A/S, 837 F.3d 1249, 1252-53 (Fed. Cir. 2016). Worse for the patentee, the very act of sending notice letters into a forum can supply the minimum contacts for personal jurisdiction over the patentee there. See Trimble Inc. v. PerDiemCo LLC, 997 F.3d 1147 (Fed. Cir. 2021); Jack Henry & Associates, Inc. v. Plano Encryption Technologies LLC, 910 F.3d 1199, 1205-06 (Fed. Cir. 2018).

The result is a genuine strategic bind, and it has no one-size answer. A patentee who marked from day one never faces it — constructive notice is already running, so there is no need to send a charge-of-infringement letter merely to start the clock, and the patentee can choose its own moment and forum to sue. A patentee who failed to mark must weigh the value of the pre-notice damages it can capture by sending a letter against the risk of forfeiting forum choice and being haled into a hostile venue. One important point of comfort cuts the patentee's way: simply giving good-faith notice of patent rights is protected, and a competitor cannot enjoin it absent a showing that the patentee's infringement position is objectively baseless. See Myco Industries, Inc. v. BlephEx, LLC, 955 F.3d 1, 11 (Fed. Cir. 2020). Still, the safest course remains the obvious one: mark, so that the letter is a tactical option rather than a forced move.

A Worked Example: The Cost of Not Marking

Consider Acme Tooling, Inc. (hypothetical), which holds a utility patent on a self-aligning drill chuck and has sold the patented chuck since January 2019 without ever marking it. Beta Hardware Co. (hypothetical) began selling an identical chuck in January 2020. In June 2026, Acme sues. If Acme had marked from 2019, the six-year window of § 286 would let it recover for essentially all of Beta's infringing sales (Beta started in 2020, inside the window), and after SCA Hygiene Beta could not invoke laches to trim that recovery for Acme's delay. Because Acme failed to mark, it recovers nothing for the period before it gave actual notice. If Acme's first specific charge of infringement was the complaint served in June 2026, Acme has forfeited damages on roughly six and a half years of infringing sales. Had Acme instead sent a proper notice letter in March 2022 naming the patent and the accused chuck, it could recover from March 2022 forward — better, but still a multi-year gap, and it would have had to weigh the declaratory-judgment risk before sending. The lesson is stark: marking is not a formality, it is often the largest single lever on the size of a damages award. Note one thing the gap does not change: even though Acme's recoverable period shrinks, the date of the hypothetical negotiation for setting a reasonable royalty remains the date of first infringement, not the date of notice. LaserDynamics, Inc. v. Quanta Computer, Inc., 694 F.3d 51, 75 (Fed. Cir. 2012). The marking limitation cuts off when the meter runs, not the rate at which it would have run.

How to Mark Correctly

The content requirements are simple but unforgiving. The mark must include the word "patent" or the abbreviation "pat." and the number of the patent. A bare "Patent Pending" or a generic "Patented" without a number does not satisfy § 287; the number must appear. The mark must be associated with the article in a way the public can perceive — courts require that the marking be sufficiently visible and durable. See Arcadia Machine & Tool, Inc. v. Sturm, Ruger & Co., 786 F.2d 1124 (Fed. Cir. 1986).

The preferred location is the article itself. When the character of the article makes marking it impracticable — because it is too small, would be defaced, or cannot physically bear the mark — the statute permits marking the package instead. Courts evaluate the practical limitations honestly rather than mechanically; the Supreme Court accepted package marking where article marking was not feasible well over a century ago. Sessions v. Romadka, 145 U.S. 29, 50 (1892).

Two requirements that clients underestimate deserve emphasis. First, marking must be substantially consistent and continuous. A patentee who marks some units and not others, or who marks for a period and then stops, has not given constructive notice during the unmarked stretches. The Federal Circuit has held that "once marking has begun, it must be substantially consistent and continuous in order for the party to avail itself of the constructive notice provisions of the statute." Nike, 138 F.3d at 1446. Sporadic marking is, for damages purposes, close to no marking at all. Second, the burden of proving compliance is on the patentee. It is the patentee who must come forward with evidence that the marking statute was satisfied. Maxwell v. J. Baker, Inc., 86 F.3d 1098, 1111 (Fed. Cir. 1996). We return below to a recent and important wrinkle on how that burden interacts with the accused infringer's pleading obligation when licensees are involved.

Virtual Marking: The America Invents Act's Website Option

For most of the statute's history, marking meant physically printing patent numbers on products or packages — a real burden for companies with deep portfolios, products covered by many patents, or patents that issue and expire on a rolling basis. Every time a patent issued or expired, the molds, labels, or engravings had to change. The Leahy-Smith America Invents Act (AIA) of 2011 solved this with virtual marking.

As amended, § 287(a) now permits a patentee to satisfy the marking requirement by fixing to the article the word "patent" or "pat." together with an internet address of a posting on the publicly accessible internet that associates the patented article with the number of the patent. In other words, the physical product carries a short, stable phrase plus a URL — for example, "Patented. See www.example.com/patents" — and the referenced web page lists which products are covered by which patent numbers. The web page does the work that once had to be done on the product surface.

Virtual marking is sometimes called the "Nike option" because consumer-products companies, including Nike, were early and prominent adopters; the approach suits footwear, apparel, electronics, and other goods with many SKUs and overlapping patent coverage. The advantages are considerable:

  • Lower cost and friction. When a patent issues or expires, the company updates a web page rather than retooling molds, reprinting packaging, or re-engraving parts.
  • Capacity. A small product surface that could never bear a dozen patent numbers can carry one URL pointing to a page that lists all of them.
  • Accuracy over time. Because the page is the source of truth, the company can keep it current and document its changes, which both supports constructive notice and reduces false-marking exposure.

Virtual marking comes with its own discipline. The URL must point to a publicly accessible page — no login walls — and the page must actually associate the patented article with the patent number, meaning a reader can tell which specific products are covered by which specific patents. A generic page that simply lists a company's patents without tying them to products is risky, because the statute requires the association. The page must be maintained: if a listed patent expires or a product is discontinued, the page should be updated promptly, both to preserve constructive notice and to avoid the false-marking problems discussed below. Prudent companies also archive dated snapshots of the page — through periodic captures or a third-party archiving service — so that, years later in litigation, they can prove what the page said on any given day. Constructive notice depends on what a member of the public would have found at the relevant time, which means the evidentiary problem is not just maintaining the page but proving its historical contents. On authenticating such web evidence in court, see Capturing the Web -- A Practitioner's Guide to Authenticating Website Screenshots as Evidence in Federal Court.

Worked Example: Virtual Marking Done Right

Crestline Audio, Inc. (hypothetical) sells a line of wireless earbuds covered by four utility patents and two design patents, with two applications pending and one patent set to expire next year. Printing eight numbers on a pea-sized earbud is impossible, and reprinting tiny packaging every time a patent issues or expires is wasteful. Crestline instead molds "Pat. crestlineaudio.com/ip" onto the charging case and prints the same on the box. At crestlineaudio.com/ip it maintains a table: the "Aria X2 earbuds" row lists the four utility patents and two design patents by number; a separate column notes pending applications generically as "patent pending." When the expiring patent lapses, Crestline removes it from the table on the expiration date and saves a timestamped archive of the page. Crestline has achieved constructive notice for the entire covered line at trivial marginal cost, and it has a clean, datable record to prove it.

Marking by Licensees and the Patentee's Duty to Ensure Compliance

The marking requirement does not apply only to the patentee's own products. By its terms, § 287(a) reaches "persons making, offering for sale, or selling within the United States any patented article for or under" the patentee — which squarely includes licensees. If a patentee licenses others to make and sell the patented article, those licensees' products must be marked too. An unmarked licensee product is, for constructive-notice purposes, just as damaging as an unmarked product of the patentee itself.

This creates a real burden, because a patentee often has limited control over how a licensee actually labels its goods. The Federal Circuit has confronted the problem directly and held that the patentee bears the responsibility, but with a calibrated standard. When third parties are involved, courts ask whether the patentee made "reasonable efforts to ensure compliance with the marking requirements." Maxwell, 86 F.3d at 1111-12. A patentee that grants a license, contractually requires marking, and takes reasonable steps to monitor and enforce that requirement stands in a far stronger position than one that licenses broadly and never checks. The standard is one of reasonable diligence, not strict liability — but reasonable diligence has to be real and, ideally, documented.

Arctic Cat v. Bombardier: The Burden, and What Happens When Marking Stops

The most consequential recent decision on licensee marking is Arctic Cat Inc. v. Bombardier Recreational Products, Inc., 950 F.3d 860 (Fed. Cir. 2020) (Arctic Cat II), which followed the court's earlier opinion in the same dispute, Arctic Cat Inc. v. Bombardier Recreational Products, Inc., 876 F.3d 1350 (Fed. Cir. 2017) (Arctic Cat I). The case arose from patents on personal watercraft steering systems. Arctic Cat had licensed the patents to a third party, and the questions were whether the licensed products had been marked — and, more pointedly, who had to prove what.

The Federal Circuit confirmed two principles every patent owner and litigator should internalize. First, on the burden of production and proof: although the ultimate burden of persuasion on marking rests with the patentee, an accused infringer who wants to limit damages based on failure to mark bears an initial burden of production — it must point to specific products it believes were sold unmarked and that it contends practice the patent. Arctic Cat I, 876 F.3d at 1366-68. This is a "low bar"; the accused infringer need only identify the products, not prove they practice the patent. Once it does, the burden shifts back to the patentee to prove that the identified products either do not practice the patent or were in fact marked. This framework prevents the patentee from being forced to prove a sweeping negative about every product ever sold by every licensee, while still placing the genuine proof obligation on the patentee once specific unmarked products are identified.

Second, and most striking, Arctic Cat II addressed the cessation of marking. The accused infringer contended that the licensee had at one point sold unmarked covered products and then stopped selling them. Arctic Cat argued that because the unmarked sales had ceased before the relevant infringement period, marking no longer mattered. The Federal Circuit rejected that argument. It held that once a patentee or its licensee sells unmarked patented articles, the failure to mark is not cured merely by ceasing to sell the unmarked articles; the patentee cannot recover pre-notice damages for the period before it either began marking consistently or gave actual notice. Arctic Cat II, 950 F.3d at 864-65. The takeaway is sobering: a single historical lapse in licensee marking can foreclose constructive notice for the affected patent unless the patentee fixes the marking going forward, and even then the gap period is lost. You cannot launder a marking failure by discontinuing the offending product.

Arctic Cat therefore teaches two operational lessons. First, write marking into every license and treat it as a covenant you will actually enforce, not boilerplate. Second, understand that damaging marking history cannot be erased by abandonment — you fix a marking problem by marking properly going forward and, where a gap already exists, by giving prompt actual notice to capture damages from that date forward.

For drafting marking covenants and audit rights into license agreements, and for thinking through how marking obligations interact with field-of-use and royalty terms, see How to License Your Patent -- From Valuation to Term Sheet.

Worked Example: A Licensee Marking Gap

Northwind Pumps, LLC (hypothetical) owns a patent on a cavitation-resistant impeller and licenses Summit Fluidics, Inc. (hypothetical) to manufacture and sell pumps using it. The license says nothing about marking. For two years, Summit sells thousands of unmarked covered pumps, then redesigns its packaging and begins marking. Northwind later sues Rival Pump Corp. (hypothetical) for infringement and seeks six years of damages. Rival meets its low burden of production by identifying Summit's unmarked pumps as covered products sold without marking. Under Arctic Cat, the burden shifts to Northwind to prove either that those pumps did not practice the patent or that they were marked — and it can prove neither. Because Summit's unmarked sales destroyed constructive notice for that period and Summit's later marking does not retroactively cure the gap, Northwind's pre-notice damages are barred; it can recover only from the date it gave Rival actual notice. A one-sentence marking covenant and an annual audit would have prevented the entire loss.

Method and Process Claims: The Marking Statute Often Does Not Apply

A crucial structural point that surprises many clients is that the marking requirement frequently does not apply to method or process claims. The reason is logical: § 287 keys the marking obligation to a tangible "patented article" that can bear a mark. A pure method — a series of steps — produces no physical article on which a number could be stamped. The Federal Circuit has accordingly held that "the law is clear that the notice provisions of § 287 do not apply where the patent is directed to a process or method." American Medical Systems, Inc. v. Medical Engineering Corp., 6 F.3d 1523, 1538 (Fed. Cir. 1993).

The consequence is favorable to method-patent owners: where a patentee asserts only method claims, there is no marking requirement to satisfy, constructive notice is effectively automatic, and the patentee may recover for the full statutory period without ever having marked anything. This is one reason method claims can be quietly valuable in a portfolio — they sidestep an entire category of damages risk.

The complication arises with mixed assertions. When a patentee asserts both apparatus (product) claims and method claims, the marking requirement can bite. The Federal Circuit has held that where both apparatus and method claims of a patent are asserted, and the patentee or its licensees sold a tangible article that practices the asserted apparatus claims, the patentee must comply with § 287 to recover damages — and the failure to mark can limit damages even on the method claims. Crown Packaging Technology, Inc. v. Rexam Beverage Can Co., 559 F.3d 1308, 1316-17 (Fed. Cir. 2009); see also American Medical, 6 F.3d at 1538-39. The governing question is whether there is a physical product associated with the asserted claims that could have been marked; if there is, and it was sold unmarked, the marking limitation applies even though method claims are in the case.

The practical guidance flows directly from the doctrine. If you own a method patent and you also sell a product that practices the method, marking the product is the safe course because it preserves constructive notice no matter how the claims are later asserted. And in litigation, counsel asserting a patent with both kinds of claims should think carefully about which claims to assert: dropping the apparatus claims and pressing only the method claims can, in the right case, take the marking limitation out of the dispute entirely. We address that litigation-strategy dimension in our infringement-litigation guide linked above.

Products Covered by Multiple Patents

Modern products are often covered by several patents at once — a smartphone may implicate dozens. Two distinct issues arise.

The first is how to mark efficiently. Physically printing many numbers is often impractical, which is precisely the problem virtual marking was designed to solve: a product can carry a single URL resolving to a page that lists every patent that covers it. For companies with large portfolios, virtual marking is not merely a convenience but close to a necessity.

The second issue is accuracy, and it bridges into false marking. When you list multiple patents on a product (or on a virtual-marking page), each listed patent should actually cover the product. Listing a patent that does not cover the article — leaving an expired patent on the page, or listing a patent whose claims the product does not practice — creates false-marking exposure under § 292. The Federal Circuit's analysis in Clontech Laboratories, Inc. v. Invitrogen Corp., 406 F.3d 1347 (Fed. Cir. 2005), reflects the principle that an article marked with a patent number must be covered by at least one claim of that patent for the marking to be proper; marking a product with a patent that does not cover it is marking an "unpatented" article as to that patent. The compliance answer is straightforward in concept and demanding in practice: maintain a current, claim-checked mapping between each product and each patent you mark, and update it as patents expire or products change.

Expiration deserves special attention. A patent that has expired no longer covers anything, so continuing to mark a product with an expired number is a form of false marking (subject to the AIA safe harbor discussed below). Because expiration dates are knowable years in advance, a disciplined company schedules removals to occur on the expiration date — one of the strongest reasons to favor virtual marking, where removal is a quick edit rather than a manufacturing change.

There is also a subtler point about multi-patent marking and the consequences of an error. Marking is generally evaluated patent by patent. If a product is properly marked with Patent A but improperly marked with expired Patent B, the marking can still earn constructive notice for Patent A while creating a potential false-marking question as to Patent B. The two analyses differ: § 287 asks whether the patentee gave adequate notice of the patent it is asserting, while § 292 asks whether any of the markings were false and deceptive. A company can therefore be in good shape on the patent it wants to enforce and simultaneously exposed on a different, stale number it forgot to remove. Yet another reason to treat the product-to-patent map as a living document audited against reality rather than against last year's assumptions.

False Marking Under 35 U.S.C. § 292

The flip side of the marking regime is the prohibition on false marking. Section 292 forbids marking an unpatented article with a patent number, or using "patent pending" or similar language, with the intent to deceive the public. The provision is meant to prevent patentees from frightening off competitors with phantom or expired patent rights. Two features dominate the modern law: a demanding intent standard, and — after the AIA — a dramatically narrowed set of people who can sue.

The Elements and the Intent-to-Deceive Requirement

A false-marking claim under § 292 requires proof that (1) the defendant marked an "unpatented article" (including an article marked with a patent that does not cover it, or with an expired patent), and (2) the defendant did so with intent to deceive the public. Clontech, 406 F.3d at 1352. Intent to deceive is the linchpin, and it is hard to prove. The Federal Circuit treats it as an inquiry into whether the marker knew its representation was false: a false statement combined with knowledge of its falsity can give rise to a rebuttable presumption of intent to deceive. See Pequignot v. Solo Cup Co., 608 F.3d 1356, 1362-63 (Fed. Cir. 2010).

Critically, that presumption can be rebutted by a credible showing of good faith. In Pequignot, Solo Cup continued to mark cup lids with patent numbers after the patents expired, but did so on the advice of counsel and as part of a deliberate, cost-conscious plan to phase out the old molds over their useful life. The Federal Circuit held that Solo Cup's good-faith belief that its conduct was lawful negated the inference of intent to deceive, even though the markings were literally false. Pequignot, 608 F.3d at 1363-64. The decision establishes that mismarking expired patents is not automatically unlawful; the question is always whether the marker intended to deceive. Documented reliance on counsel, a reasonable phase-out plan, or cautious "may be covered by" language all cut strongly against deceptive intent.

The AIA Revolution: Qui Tam Abolished, Competitive Injury Required

Before 2011, § 292 was a qui tam statute — any member of the public could sue as a private attorney general "on behalf of" the United States, splitting any penalty with the government, without showing any personal injury. After the Federal Circuit held in Forest Group, Inc. v. Bon Tool Co., 590 F.3d 1295 (Fed. Cir. 2009), that the statutory penalty (up to $500) could be assessed on a per-article basis rather than per decision to mark, the math became explosive: millions of falsely marked articles meant the potential for staggering aggregate penalties. The result was a flood of opportunistic false-marking suits by plaintiffs with no connection to the products at all — a true cottage industry of bounty-hunters.

Congress shut this down in the AIA, and practitioners must know the two changes cold. First, the qui tam mechanism is abolished: only the United States may sue for the statutory penalty. Private bounty-hunter suits are gone. Second, a private party may bring a civil action only if it has suffered a competitive injury as a result of the false marking, and its recovery is limited to damages adequate to compensate for that injury — not the per-article penalty. The AIA also created a safe harbor: marking a product with the number of an expired patent that previously covered the product is not a violation of § 292. That safe harbor substantially reduces — though it does not eliminate, given other forms of mismarking — the risk that once drove the litigation wave.

The practical upshot is that false-marking litigation today is a competitor-versus-competitor affair grounded in real economic harm, not a stranger's lottery ticket. To have standing, a private plaintiff must show genuine competitive injury — generally that it competes (or made concrete, capable efforts to compete) and suffered identifiable harm causally tied to the false marking. See Sukumar v. Nautilus, Inc., 785 F.3d 1396, 1400-04 (Fed. Cir. 2015) (analyzing what it means to be a competitor capable of suffering competitive injury, including the requirement of an actual attempt to enter the market with a reasonable possibility of success); Gravelle v. Kaba Ilco Corp., 684 F. App'x 974 (Fed. Cir. 2017) (plaintiff must demonstrate harm causally connected to the false marking). The government rarely pursues false-marking penalties, so the realistic exposure for most companies is a competitor's compensatory claim — which is precisely why accurate, well-documented marking remains worthwhile even in the post-AIA era.

"Patent Pending" and Pending Applications

The phrase "patent pending" tells the public that a patent application covering the article is on file but has not yet issued. It is permissible — and useful as a deterrent — when an application is genuinely pending, and may be used as soon as an application (including a provisional application) is filed. But "patent pending" confers no enforceable rights; there is no such thing as infringing a pending application, and the phrase does not satisfy § 287's marking requirement for purposes of constructive notice of an issued patent. (It does, however, foreshadow provisional rights under 35 U.S.C. § 154(d), which can entitle a patentee to a reasonable royalty for certain pre-issuance activity once the patent issues — a distinct mechanism from § 287 marking.)

Falsely claiming "patent pending" when no application is on file can violate § 292 if done with intent to deceive, and may also support state-law unfair competition or false-advertising theories. See Third Party Verification, Inc. v. Signaturelink, Inc., 492 F. Supp. 2d 1314 (M.D. Fla. 2007). The compliance rule is simple: use "patent pending" only while an application is actually pending, and remove it promptly if the application is abandoned, finally rejected, or issues as a patent (at which point you switch to the patent number). For the relationship between false marking and competitor false-advertising claims more broadly, see False Advertising and Lanham Act Section 43(a) -- Competitor Claims in Technology Marketing.

It is also a mistake to assume that the AIA's narrowing of federal § 292 standing closed off every avenue of attack on improper marking. Although federal law governs the marking statute itself, courts have recognized that state-law unfair-competition claims premised on false or deceptive marking are not necessarily preempted, so a competitor frozen out of a federal qui tam theory may still have a state-law route grounded in the same conduct. See Technology Licensing Corp. v. Intersil Corp., No. C-08-3023, 2009 WL 5108395 (N.D. Cal. Dec. 17, 2009). The practical lesson is that a company advising on marking strategy — especially in jurisdictions with robust unfair-competition statutes such as California's — should weigh potential state-law exposure alongside the now-diminished federal § 292 risk, because the two do not rise and fall together.

Marking, Willfulness, and Enhanced Damages: Keeping the Buckets Straight

Clients frequently conflate three different "notice"-flavored concepts: marking under § 287, the willfulness inquiry that can support enhanced damages under 35 U.S.C. § 284, and pre-suit notice letters. They are related but distinct, and confusing them leads to expensive mistakes.

Marking governs when compensatory damages begin to accrue. Willfulness — whether the infringer's conduct was egregious enough to justify enhancing damages up to treble — is a separate inquiry that turns on the infringer's culpable state of mind. Halo Electronics, Inc. v. Pulse Electronics, Inc., 579 U.S. 93 (2016) (rejecting the rigid two-part Seagate test and committing enhancement to the district court's discretion for the "egregious" case). A patentee can satisfy marking and collect full compensatory damages without ever proving willfulness; conversely, a defendant's knowledge of the patent that bears on willfulness does not by itself create the "actual notice" that § 287 requires when a product was unmarked. The two doctrines even point in opposite directions on the relevance of knowledge: under § 287 the infringer's knowledge is irrelevant to the unmarked patentee's pre-notice damages, while under Halo the infringer's knowledge is central to enhancement. Keep the buckets separate: marking is about the damages window; willfulness is about the damages multiplier. A well-drafted notice letter can serve both functions at once — starting the § 287 clock and creating the knowledge that, if the infringer presses on, may later support willfulness — but it does so under two different legal theories, and counsel should understand which box each effect lands in.

Building a Patent Marking Compliance Program

Because marking errors are usually errors of omission and drift rather than deliberate misconduct, the antidote is a standing program with clear ownership, scheduled reviews, and an audit trail. The framework below is written to be implemented by a company of any size; scale the formality to the portfolio.

1. Assign ownership. Designate a responsible person or small committee — typically in-house IP counsel working with operations and marketing — who owns the marking decisions and the documentation. Marking fails most often when it is everyone's job and therefore no one's.

2. Maintain a product-to-patent map. Keep a living table linking each product (by SKU or model) to each patent that covers it, with the basis (which claims read on the product), the issue date, and the expiration date. This map is the single source of truth for both physical and virtual marking, and it is the document you will produce in litigation to prove compliance — and, under Arctic Cat, to rebut an accused infringer's identification of supposedly unmarked products.

3. Choose physical or virtual marking deliberately, and prefer virtual where the portfolio is deep or products are small. If you adopt virtual marking, host a public, no-login page that associates each product with its patent numbers, keep it current, and archive dated snapshots so you can later prove what it said on any date.

4. Mark from first sale, and mark consistently. Build marking into the product-launch checklist so the first commercial unit is marked. Because constructive notice requires substantially continuous marking, never let marking lapse and resume; a gap can cost the damages for that entire period.

5. Bind your licensees. Every license should require the licensee to mark covered products in compliance with § 287, give you audit rights, and obligate prompt cure of any lapse. Then actually exercise the audit rights. Arctic Cat makes clear that licensee lapses are your problem, so make licensee compliance contractual and monitored, not aspirational.

6. Run a calendared expiration and abandonment sweep. Schedule removals of patent numbers to occur on expiration dates and removals of "patent pending" on abandonment or issuance. Continuing to mark expired patents or phantom pending applications is the classic path to false-marking exposure, even with the AIA's expired-patent safe harbor.

7. Document good-faith decisions. Where a marking question is genuinely close, memorialize the reasoning — and, for higher-stakes calls, obtain an opinion of counsel. Pequignot shows that documented good faith is a powerful shield against any inference of intent to deceive.

8. Audit periodically. At least annually, reconcile the product-to-patent map against what is actually being marked on products, packaging, and the virtual-marking page, and against the current status of every listed patent. Keep the audit reports; they are evidence of reasonable efforts.

9. Coordinate marking with notice and enforcement strategy. If you discover an unmarked period that cannot be cured, recognize that your route to pre-notice damages for that gap is actual notice, and consider sending a properly drafted charge of infringement to begin the clock — while weighing the declaratory-judgment and personal-jurisdiction risks that such letters create. This is where marking compliance hands off to litigation strategy, covered in our Comprehensive Guide to Patent Infringement Litigation.

A program along these lines costs little and converts marking from a litigation surprise into a managed, provable process.

Key Takeaways

Patent marking is a high-leverage, low-cost discipline that too many patent owners treat as an afterthought until a damages expert delivers bad news. The essentials: marking is not mandatory, but under § 287 the failure to mark bars all pre-notice damages except where the patentee proves actual notice — a specific charge of infringement against a specific product — so marking is usually the difference between recovering for six years of infringement and recovering only for the months after a lawsuit is filed. With laches abolished by SCA Hygiene, marking is now the principal limit on the reach of a damages claim, which makes getting it right more important than ever. Constructive notice, earned by substantially continuous marking, is the prize because it makes the infringer's knowledge irrelevant and unlocks the full § 286 window. The AIA's virtual-marking option lets companies mark with a URL — essential for small products and deep portfolios — but demands a current, well-documented, publicly accessible page that ties products to patents, with archived snapshots to prove its historical contents. Licensee products must be marked too, and Arctic Cat confirms both the burden-shifting framework and the hard truth that a marking gap cannot be cured by ceasing the unmarked sales — so marking covenants and audits belong in every license. Method-only claims generally escape the marking requirement, but mixed assertions can pull the limitation back in. And false marking under § 292, far less dangerous after the AIA abolished qui tam and required competitive injury, still warrants accurate, documented marking, especially as to expired patents and "patent pending."

For the broader picture that situates patent marking alongside trademark and trade dress marking, see the companion guide, Understanding Patent and Trade Dress Marking Requirements -- A Comprehensive Guide for Businesses and Practitioners. For a plain-English grounding in the patent system generally, see Patent Basics -- A Plain English Guide and Patent FAQs -- Answers to Common USPTO Patent Questions.

Frequently Asked Questions

Is patent marking required by law? No. Marking is not a condition of patent validity, and a patentee can sue for infringement whether or not it marked. But marking determines when damages start: without marking (or actual notice), § 287 bars recovery for the entire pre-notice period. So while marking is optional, failing to mark can forfeit years of damages — and after SCA Hygiene abolished laches, marking is the main thing standing between a patentee and the full six-year damages window.

What is the difference between constructive notice and actual notice? Constructive notice is achieved by properly and consistently marking the patented article; it puts the whole world on notice as a matter of law, regardless of whether any infringer actually saw the mark, and it lets damages run for the full statutory period. Actual notice is an affirmative communication from the patentee charging a specific product with infringing a specific patent; it starts damages only from the date of that communication forward. Marking is far more valuable because it captures pre-notice damages and never depends on the infringer's knowledge.

Can I mark my products with just a website instead of patent numbers? Yes, under the America Invents Act's virtual-marking provision. You mark the article with "patent" or "pat." plus a URL to a publicly accessible page that associates the patented article with its patent numbers. The page must be free of login walls, must actually tie products to patents, and should be kept current and archived so you can prove what it said on any given date.

Do I have to make sure my licensees mark their products? Yes. The marking requirement extends to licensees, and under Arctic Cat Inc. v. Bombardier Recreational Products, Inc., 950 F.3d 860 (Fed. Cir. 2020), an unmarked licensee product can destroy your constructive notice. You must make reasonable efforts to ensure licensee compliance — typically a marking covenant in the license plus audit rights you actually use. A licensee's marking lapse cannot be cured simply by discontinuing the unmarked product.

Do method or process patents need to be marked? Generally no. Because a pure method produces no physical article to mark, § 287's marking requirement does not apply to method-only assertions, and damages run for the full period without any marking. But if you assert both apparatus and method claims and you sold a tangible product practicing the apparatus claims, the marking requirement can apply and limit damages even on the method claims — so marking the product is the safe course.

If I never marked, can I still send a letter to start collecting damages? Yes — that is the actual-notice route. A letter that identifies the patent by number, identifies the specific accused product, and charges that it infringes will start your damages clock from the date of receipt. The catch is that the same specificity can give the recipient grounds to file a declaratory-judgment action in its preferred court under MedImmune, Inc. v. Genentech, Inc., 549 U.S. 118 (2007), and may even establish personal jurisdiction over you there. Weigh the pre-notice damages you can capture against the loss of forum control — and remember that if you had simply marked, you would not face this trade-off at all.

How risky is false marking after the America Invents Act? Much less risky than before. The AIA abolished qui tam false-marking suits, so only the U.S. government can recover the statutory penalty, and private plaintiffs must prove actual competitive injury and recover only compensatory damages. The AIA also created a safe harbor for marking with an expired patent that once covered the product. Intent to deceive remains essential and is hard to prove, especially against documented good faith. Still, accurate marking — promptly removing expired patents and phantom "patent pending" claims — remains the right practice.

Related Articles

This article is provided by mclaw.io for general informational purposes only and does not constitute legal advice, nor does it create an attorney-client relationship. Patent marking and false-marking law are fact-specific and evolving; the timing of damages and the consequences of a marking lapse can turn on details unique to your products, licenses, and litigation posture. For guidance on your particular situation, please consult qualified patent counsel.