Pick up almost any product within arm's reach and turn it over. Somewhere on the housing, the box, or a buried "about" screen, there is probably a string of fine print: a patent number, a circled R, a copyright line, maybe a URL you would never visit voluntarily. To most people that fine print is visual noise. To an intellectual-property lawyer it is a confession — a statement, made under several different statutes at once, about exactly which legal rights the manufacturer claims and how much money it expects to be able to collect if someone copies the thing.

That is what marking is: the public-notice machinery of intellectual property. A single consumer product can be wrapped in four or five different rights simultaneously. The mechanism inside may be covered by a utility patent. Its distinctive shape may be covered by a design patent and also function as protectable trade dress. The brand name on the box is a trademark. The packaging artwork, the printed manual, and the software user interface are copyrighted. Each of those rights has its own marking or notice regime — and here is the part that surprises even experienced businesspeople — those regimes do not work the same way. Some marking is mandatory in the practical sense that skipping it forfeits money. Some is purely optional but strategically valuable. And in at least one corner, marking something you should not mark is itself a tort.

This article is the broad, comparative companion to our focused, practitioner-level treatment of patent marking. For a deep, step-by-step walkthrough of the patent marking statute, the mechanics of a virtual-marking page, licensee compliance, and the case law on actual notice, see Understanding Patent Marking Requirements -- A Practical Guide for Practitioners and Entrepreneurs. The goal here is different: to set all of the major U.S. IP marking and notice regimes alongside one another, explain how each one affects the rights holder's money and remedies, flag the traps unique to each, and then fuse them into a single coordinated compliance program for a product that carries several rights at once.

The audience is deliberately broad. A business owner deciding what to print on a label should be able to follow it. So should a lawyer drafting a marking policy or litigating a damages cut-off, and so should a judge who wants the framework laid out cleanly. Every term of art is explained in plain language the first time it appears, and the discussion is anchored throughout in primary authority — the statutes and the leading cases interpreting them.

Why Marking Matters: Notice, Money, and the Public

Before diving into the individual regimes, it helps to understand what marking is for, because the purpose explains the rules.

All of these regimes share one underlying idea: notice to the public. Intellectual property rights are intangible. You cannot see a patent claim or a trademark registration by looking at a product; the right exists on paper in a government office, not on the shelf. Marking is the device the law uses to bridge that gap — to tell the world, "this thing is protected, and here is roughly how." That notice warns would-be copyists, helps the marketplace identify protected goods, and reduces inadvertent infringement. The law would rather you put the world on notice than ambush an unsuspecting copier with a damages claim it had no way to anticipate.

Where the regimes diverge sharply is in the legal consequence of giving or withholding that notice. Three distinct models operate across the IP landscape, and almost everything that follows is a variation on one of the three.

The first model is damages-gating: notice is the price of admission to a particular pool of damages. Patent marking works this way. You are never required to mark, but if you sell a product covered by a patent and you do not mark it, you generally cannot collect damages for infringement that occurred before you gave the infringer actual notice. The notice does not create the right; it unlocks back-damages.

The second model is profit- and damages-conditioning with a fault overlay. Federal trademark and copyright both have versions of this. For registered marks, the registration symbol affects whether a defendant who pleads ignorance can be made to pay profits and damages for the period before notice. For copyright, the optional notice defeats a defendant's "innocent infringement" plea and thereby protects the full size of the damages award.

The third model is prohibition: certain marks may not be used unless the corresponding right actually exists. You cannot legally place the federal registration symbol on an unregistered mark, and you cannot mark a product with a patent number that does not cover it (or that has expired) with intent to deceive. Here, marking is not a privilege you forfeit by inaction but a representation you can be punished for making falsely.

Hold those three models in mind and the rest of the article snaps into focus. As we walk through patents, trademarks and trade dress, and copyright, three questions recur: Is marking mandatory or optional? What does it unlock or protect? And what is the penalty for getting it wrong in either direction — too little or too much?

Patent Marking Under 35 U.S.C. § 287

The basic rule and what it unlocks

Patent marking is governed by 35 U.S.C. § 287(a). The statute is permissive in form but punishing in effect. It says that a patentee, and anyone making or selling a patented article on the patentee's behalf, "may give notice to the public that the same is patented" by marking the article with the word "patent" or the abbreviation "pat." together with the patent number. Then comes the sentence that does all the work: "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."

Translated: you do not have to mark, but if you sell physical products embodying your patent and you do not mark them, you lose the right to collect damages for any infringement that happened before you gave the infringer actual notice. Marking the product instead gives the world constructive notice — the law treats everyone as on notice from the date the marked products went into the market, so damages can run from the start of the infringement. The difference is purely about money, and in a long-running infringement it can be the difference between a seven-figure recovery and a token one.

The Supreme Court fixed the contours of this rule more than a century ago in Dunlap v. Schofield, 152 U.S. 244 (1894), holding that a patentee who fails to mark must prove the infringer was actually notified, and that the burden of proving compliance rests on the patentee. Modern Federal Circuit law confirms both points. In Arctic Cat Inc. v. Bombardier Recreational Products Inc., 950 F.3d 860 (Fed. Cir. 2020), the court held that although an accused infringer bears an initial burden of production to identify unmarked products it believes practice the patent, the ultimate burden of persuasion — proving the products were properly marked, or that marking was excused — stays with the patentee throughout. Arctic Cat drove the point home with an unforgiving holding on the cut-off itself: once the marking obligation is triggered and breached, the damages bar persists until actual notice is given, and it is not cured merely because the patentee later stops selling the unmarked product. Marking compliance, in other words, is not a box you check once; it is a status you must maintain.

A critical limit: marking only matters when there is a product to mark

A point that confuses many businesspeople is that the marking requirement attaches only when the patentee (or a licensee) actually makes, offers, or sells a tangible article that embodies the patent. If the patent claims only a method or process — something with no physical article to stamp — there is nothing to mark, and the marking statute does not limit damages at all. The Federal Circuit confirmed this in American Medical Systems, Inc. v. Medical Engineering Corp., 6 F.3d 1523 (Fed. Cir. 1993), and reaffirmed the principle in Crown Packaging Technology, Inc. v. Rexam Beverage Can Co., 559 F.3d 1308 (Fed. Cir. 2009), which examined how the rule applies when a patent contains both apparatus and method claims and only the method claims are asserted. Likewise, a patentee that licenses or sells nothing embodying the patent — the classic non-practicing entity, or "patent troll" — has no products to mark and is therefore not bound by § 287's damages limitation. The marking obligation is triggered by selling a product, not by owning a patent. (For the threshold question of whether a given product even falls within a patent's claims in the first place, see What Constitutes Patent Infringement and our explainer on Patent Basics.)

"Substantially all" and the consistency trap

When the patentee does sell embodying products, the marking must be substantially consistent. Courts require marking of "substantially all" patented articles, and they require it to be continuous once begun. Sporadic or partial marking can defeat constructive notice for the entire product line. This is one of the most common and costly errors in patent practice: a company marks most of its units, a meaningful fraction goes out unmarked, and a court holds that no constructive notice attached at all — forcing the patentee all the way back to the actual-notice date and erasing years of damages.

The problem is amplified by licensees. Because § 287 reaches articles sold "by or under [the patentee]," a licensee's unmarked sales can destroy the patentee's constructive notice just as surely as the patentee's own. The Federal Circuit's decision in Maxwell v. J. Baker, Inc., 86 F.3d 1098 (Fed. Cir. 1996), recognized a "rule of reason" that can excuse a patentee who made reasonable efforts to ensure licensee marking, but no prudent rights holder leans on that grace. The safer reading is that patentees must contractually require marking and actively monitor compliance. License agreements should contain an express marking covenant, audit rights, and a mechanism to update patent information when the portfolio changes — the same discipline we recommend in How to License Your Patent -- From Valuation to Term Sheet.

Actual notice: what counts and what does not

If a product was not marked, damages start only when the infringer received actual notice — and the Federal Circuit sets a demanding standard for what qualifies. Under Amsted Industries Inc. v. Buckeye Steel Castings Co., 24 F.3d 178 (Fed. Cir. 1994), actual notice requires an affirmative communication from the patentee to the accused infringer that identifies a specific patent and a specific accused product or activity and charges infringement. A vague statement that the patentee "has patents," or a polite reminder to "be careful," is not enough. Critically, actual notice is measured by the patentee's conduct, not the infringer's knowledge: the fact that the infringer independently knew the patent existed does not start the damages clock. Filing the infringement complaint always counts as actual notice — but by then most of the back-damages window may already be gone. The lesson for enforcement is to send a notice letter that names the patent and the product before reaching for the complaint, a topic that overlaps with the strategy in Drafting a Trademark Cease and Desist Letter (the patent analogue follows the same logic).

Virtual marking

The America Invents Act of 2011 added a modern alternative. Under § 287(a) as amended, a patentee may satisfy the marking requirement by virtual marking — placing the word "patent" or "pat." together with a freely accessible internet address on the product, where that web page associates the patented article with its patent number(s). Virtual marking solves the operational nightmare of re-tooling molds and re-printing labels every time a patent issues, expires, or is invalidated. The patentee simply updates the web page; the product itself bears only a neutral URL.

Virtual marking carries its own discipline. The web address must appear on the article (or its packaging where the article cannot practically be marked); the page must be publicly accessible without a paywall or login; and it must accurately associate specific products with specific patents. A page that is frequently offline, that lists patents without tying them to products, or that is out of date can fail to provide constructive notice — and, as discussed below, listing expired or non-covering patents can edge toward false marking. The mechanics, including how courts evaluate the adequacy of a virtual-marking page, are treated in depth in our practical patent marking guide.

Design patents and marking

Design patents — which protect the ornamental appearance of an article rather than its function — are also subject to § 287. A design patent number can and should be marked on (or virtually marked for) the article whose appearance it protects. Because design patents have become central to consumer-product protection (and because design and utility patents frequently cover the very same gadget — an overlap explored in The Intersection of Design and Utility Patents), a complete marking program lists both. The conventional format reads "U.S. Patent No. D[number]" for a design patent, the leading "D" making the design-versus-utility distinction transparent. Design patents are also where the patent-versus-trade-dress overlap becomes most acute, a relationship we return to below and dissect in The Evolution of Design Patent Functionality.

False marking under 35 U.S.C. § 292

Patent marking is bounded on the other side by the false marking statute, 35 U.S.C. § 292. It prohibits marking an article with a patent number (or with words like "patent pending") when the article is not in fact covered, or when the patent has expired — if done with intent to deceive the public. Two developments, one judicial and one legislative, reshaped this area and are essential to understand.

The judicial spark was Forest Group, Inc. v. Bon Tool Co., 590 F.3d 1295 (Fed. Cir. 2009). The false-marking statute then provided a fine of "not more than $500" for the offense, and for decades courts had read that fine as a single penalty per decision to mark — a trivial sum. Forest Group held instead that the penalty accrues per falsely marked article. Suddenly a single mismarked patent number on a high-volume product implied a theoretical penalty running into the millions. Because the pre-2011 statute was a qui tam provision — any member of the public could sue and keep half the penalty, with the other half going to the government — Forest Group detonated a litigation boom. Opportunistic plaintiffs, sometimes shell entities created for the purpose, fanned out across hardware stores buying products marked with long-expired patents and filing hundreds of "marking troll" suits. (The Federal Circuit tried to cabin the damage in Pequignot v. Solo Cup Co., 608 F.3d 1356 (Fed. Cir. 2010), holding that marking products with expired patents is "false," but that there is a rebuttable presumption against deceptive intent where a company acts in good-faith reliance on counsel; a high-volume marker, in other words, was not automatically liable.)

Congress responded the following year in the America Invents Act, and it responded decisively. The AIA eliminated the qui tam bounty entirely. Now only the United States may sue for the statutory per-article penalty, and a private plaintiff may sue only if it has suffered a competitive injury from the false marking — and may recover only damages adequate to compensate for that injury. (Pub. L. No. 112-29, § 16, 125 Stat. 284, 329 (2011), amending 35 U.S.C. § 292; the changes applied to all cases pending on or commenced after September 16, 2011.) That single change drained the swamp; the marking-troll docket largely evaporated within a year.

The AIA's second move was equally practical: it added 35 U.S.C. § 292(c), which provides that marking a product with an expired patent that previously covered the product is not a violation. This safe harbor matters enormously for physical marking, because it means a company need not frantically re-tool molds the moment a patent expires. It does not, however, license affirmative deception, and the intent-to-deceive element remains for patents that never covered the product. The cleanest way to avoid false-marking exposure altogether — expired patents, non-covering patents, and stale lists alike — is virtual marking with a diligently maintained page.

The takeaway for the comparative picture is striking: patent marking is the regime where both under-marking and over-marking carry consequences. Under-mark, and you forfeit back-damages under § 287. Over-mark with deceptive intent, and you risk a false-marking claim under § 292. The well-run program threads the needle by marking accurately, consistently, and verifiably. There is no parallel double-edged exposure in copyright, and only a faint one in trademark.

Trademark and Trade Dress Registration Notice Under 15 U.S.C. § 1111

What trade dress is, and why marking it is unusual

A trademark is any word, name, symbol, or device that identifies the source of goods or services (15 U.S.C. § 1127). Trade dress is a particular flavor of trademark: the overall look and feel of a product or its packaging — shape, color, texture, configuration, layout — that signals source. The classic examples are the contour shape of a Coca-Cola bottle and the décor of a restaurant chain. The Supreme Court has confirmed that trade dress is a "symbol or device" within the Lanham Act's definition of a trademark (see Wal-Mart Stores, Inc. v. Samara Bros., 529 U.S. 205, 209–10 (2000)), and it is protected under both Section 32 for registered trade dress (15 U.S.C. § 1114) and Section 43(a) for registered and unregistered trade dress (15 U.S.C. § 1125(a)). Trade dress sits at a fascinating intersection of branding and design, with doctrinal complications — functionality, distinctiveness, secondary meaning — that we examine in depth in The Intricate World of Trade Dress Protection and in Design Patents vs. Trade Dress Protection. For the foundations of trademark law generally, see Trademark Basics.

For marking purposes, the crucial fact is that trade dress is rarely something you can stamp with a symbol in any natural way. You cannot easily put an ® on the shape of a bottle. As a result, trade dress notice is usually given through a footnote, hangtag, packaging legend, or website statement rather than a symbol on the article itself. But the legal rules that govern trademark notice apply to trade dress in exactly the same way, because trade dress is a species of trademark. So the symbol rules below apply, with the practical caveat that the medium of notice is often packaging or marketing materials.

The federal registration symbol and what it conditions

Federal trademark notice is governed by 15 U.S.C. § 1111. Like the patent statute, it is permissive in form: the owner of a federally registered mark may give notice of registration by displaying the federal registration symbol — the encircled R, ® — or one of two equivalent statutory phrases, "Registered in U.S. Patent and Trademark Office" or its abbreviation "Reg. U.S. Pat. & Tm. Off."

What that notice conditions is the recovery of profits and damages in an infringement suit. Section 1111 provides that, in a suit for infringement of a registered mark, a plaintiff who did not display registration notice may not recover profits and damages "unless the defendant had actual notice of the registration." In plain terms: if you own a federal registration but never marked your goods with ® (or the equivalent phrase), and you cannot prove the defendant actually knew about your registration, you forfeit the ability to collect the defendant's profits and your damages for the unmarked period. Injunctive relief is still available, but the money is gone. This mirrors the patent constructive-notice idea: displaying ® gives constructive notice of the registration, and that notice unlocks the monetary remedies.

It is worth being precise about what those monetary remedies are, because § 1111 is the gatekeeper to a generous remedial scheme. The Lanham Act authorizes recovery of (1) the defendant's profits, (2) the plaintiff's actual damages, which a court may in its discretion enhance up to three times the actual amount in appropriate cases, (3) attorneys' fees in exceptional cases, and (4) destruction of infringing articles (15 U.S.C. §§ 1117, 1118). Section 1111 stands between the unmarked plaintiff and the first two of those — the profits and damages — for the period before notice. That is a lot of money to leave on the table for want of a symbol that costs nothing to type.

There is an important difference from patent law worth flagging. Patent marking gates all pre-notice damages; § 1111 is keyed specifically to the registration, addressing profits and damages for the period before the defendant had notice of the registration. And courts read the actual-notice escape hatch somewhat generously to plaintiffs — a defendant's documented knowledge of the registration can substitute for the symbol. Still, no competent brand owner litigates actual notice when displaying ® is free. The ® symbol is the cheapest insurance in all of intellectual property.

Registration is a prerequisite for ®, and improper use has real consequences

The single most important rule about the ® symbol is the one most often violated: you may use ® only with a mark that is actually the subject of a current federal registration, and only in connection with the goods or services listed in that registration. Using ® on a mark that is merely applied-for, on a mark for goods not covered by the registration, or on a mark whose registration has lapsed is improper.

Why does this matter? Improper use of ® can carry several adverse consequences. The USPTO can treat a fraudulent or improper claim of registration as a defect that taints an application; more dangerously, courts have held that a knowing misuse of the registration symbol — using ® to deceive the public or competitors into believing a mark is registered when it is not — can constitute unclean hands and bar relief that would otherwise lie under § 1114 or § 1125(a). The settled articulation is that misuse of the symbol bars relief only when it is deliberate and intended to deceive or mislead, and not where it results from inadvertence, ignorance, or a good-faith but mistaken belief that registration had issued. But "we didn't mean it" is a fragile defense to build a brand on. The safe practice is rigid: ® goes on only after the registration certificate issues, and only for the registered goods and services. The path from application to that certificate is laid out in How to File a Trademark Application with the USPTO and the Federal Trademark Application Checklists. Timing also matters in the run-up to registration: even after a Notice of Allowance for an intent-to-use application, the mark is not yet registered, so ® must wait.

TM and SM for unregistered and pending marks

What may a business use before — or instead of — registration? The superscript ™ (trademark) and ℠ (service mark) symbols. These have no statutory definition and no statutory effect; they are creatures of custom. TM asserts common-law trademark rights in goods; SM asserts common-law rights in services. Anyone may use them on any mark they claim, registered or not, pending or not. They do not invoke § 1111 and do not by themselves unlock profits and damages.

Their value is strategic and evidentiary rather than statutory. Using TM/SM signals to the public and to competitors that the owner claims the mark as a source identifier. That signal can support the single most important element of unregistered-trademark and trade-dress protection: secondary meaning (also called "acquired distinctiveness"), the requirement that consumers have come to associate the mark or trade dress with a single source. This is not a minor point for trade dress. The Supreme Court has drawn a bright line that product design or configuration — as opposed to product packaging — can never be inherently distinctive and always requires proof of secondary meaning (Wal-Mart, 529 U.S. at 215–16), and the same is true of a single product color (Qualitex Co. v. Jacobson Products Co., 514 U.S. 159, 162–63 (1995)). Product packaging trade dress, by contrast, can be inherently distinctive (Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763 (1992)), and where the classification is ambiguous, courts err toward treating it as design and demanding secondary meaning (Wal-Mart, 529 U.S. at 215). Because so much valuable trade dress falls into the configuration bucket, any evidence that the owner consistently claimed and used the trade dress as a brand — a wall of TM-marked specimens stretching back years — is gold. (Trade dress must also be non-functional to be protectable, a separate hurdle confirmed in TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 (2001), and codified for the plaintiff's burden in § 1125(a)(3).) Consistent TM use, prominent and sustained, is a brick in the secondary-meaning wall. The same record can also help fix the geographic reach of an unregistered right: because common-law trade dress rights extend only as far as the owner's actual use, reputation, and consumer recognition, courts weigh consistent TM marking — alongside sales and advertising evidence — when fixing the territorial scope of the protectable right, an inquiry that a federal registration would otherwise short-circuit by conferring presumptively nationwide rights. Marking the goods as they enter each new market is thus part of the proof that the right traveled there too.

Two cautions. First, TM/SM is not a substitute for ® once registration issues — keep using ® on registered marks to preserve § 1111 notice. Second, while TM/SM carries no false-marking penalty analogous to § 292, claiming trademark rights you do not have can still expose you to unfair-competition or declaratory-judgment risk if you use the claim to bully competitors. Use TM/SM honestly.

Trade dress notice in practice

Because trade dress is usually un-stampable, give notice through legends. A typical packaging or website legend reads: "The [product] configuration is a registered trademark of Acme Corp." for registered trade dress, or "Acme Corp. claims trade dress rights in the configuration of the [product]" for unregistered trade dress. For litigation purposes, what matters is consistent, documented assertion of the claimed trade dress as a source identifier — both to satisfy any actual-notice question under § 1111 and to build the secondary-meaning record. This dovetails with the broader online enforcement posture discussed in Brand Protection Online -- A Strategic Guide for Businesses and with timing considerations covered in When to Trademark Your Brand.

Copyright Notice Under 17 U.S.C. §§ 401–402

Optional since 1989, but not pointless

Copyright notice is the regime that changed most dramatically in living memory, and the change is widely misunderstood. The familiar notice — the © symbol, the year of first publication, and the name of the owner — is governed by 17 U.S.C. § 401 (for visually perceptible copies) and § 402 (for phonorecords of sound recordings). For works first published in the United States on or after March 1, 1989, notice is optional. That date is when U.S. adherence to the Berne Convention took effect, and Berne forbids conditioning copyright on formalities like notice. So the old, draconian rule — that publishing without proper notice could inject a work into the public domain — is gone for post-1989 works.

(For works published before 1989, the older, formality-laden regime can still matter, because the public-domain consequences of a defective or absent notice attached at the moment of that earlier publication and may not be curable. Anyone evaluating the status of an older work should analyze it under the law in force when it was published. For modern works, the optional-notice rule controls.)

What notice still does: defeating innocent infringement

If notice is optional, why bother? Because § 401(d) and § 402(d) attach a concrete litigation benefit to it. If a proper notice appears on the published copies to which a defendant had access, then in a later infringement suit the defendant may not interpose a defense based on innocent infringement in mitigation of actual or statutory damages — except as that defense bears on the narrow reduced-award provisions of § 504(c)(2) (certain library and nonprofit uses).

Here is why that matters in dollars. Under the Copyright Act's statutory-damages regime, 17 U.S.C. § 504(c), a plaintiff who registered in time may elect statutory damages instead of proving actual loss. The ordinary range runs from $750 to $30,000 per work, rising to as much as $150,000 per work for willful infringement. But the statute also lets a court reduce an award to as little as $200 per work where the infringer proves it was an innocent infringer — that it was not aware and had no reason to believe its acts infringed. A visible © notice on the copies the defendant accessed forecloses that innocent-infringer reduction. So copyright notice does not unlock damages the way patent and trademark marking do; instead it protects the size of an available statutory-damages award by stripping the defendant of its single most common mitigation argument. (Statutory damages and attorneys' fees also require timely registration under § 412 — notice and registration are complementary, not interchangeable. The registration process is covered in How to Register a Copyright with the U.S. Copyright Office, and the strategic role of notice in Copyright Notice -- Form, Function, and Best Practices.)

Form and placement

A proper notice under § 401 has three elements: (1) the symbol © (or the word "Copyright" or the abbreviation "Copr."); (2) the year of first publication of the work; and (3) the name of the copyright owner. Example: "© 2026 Acme Corp." For sound recordings, § 402 substitutes the encircled ℗ (the "phonorecord" symbol) for ©. The notice must be affixed in a manner and location that gives reasonable notice of the claim — on the title page, in the gutter, on the packaging, in the credits, or in a footer for digital works.

Because notice is free, easy, and protective, the rule for businesses is blunt: put a proper © notice on everything you publish — manuals, packaging artwork, websites, software splash screens and "about" boxes, marketing collateral, and audiovisual works. There is no downside and a real litigation upside. Software in particular is best protected through a layered strategy — copyright, plus possibly patent and trade-secret protection — as discussed in Legal Protection of Software -- Copyrights, Patents, Trade Secrets, and Contracts.

Where the Regimes Overlap on a Single Product

One reason a comparative treatment earns its keep is that the same physical feature of a product can be protected by more than one regime at once, and each regime's marking rules then apply independently and simultaneously.

Consider the shape of a product. That shape may be the subject of a design patent — an ornamental design right that, for designs filed under current law, expires fifteen years from grant — and the same shape may also function as trade dress, a potentially perpetual source identifier so long as it remains distinctive, non-functional, and in use. The two rights are conceptually distinct: one rewards novel ornamental design for a fixed term, the other protects consumer source-association indefinitely. The relationship between them is explored at length in Design Patents vs. Trade Dress Protection for Product Configurations and in The Evolving Landscape of Product Configuration Protection.

For marking, the overlap means the owner should mark both rights, because each marking statute protects a different remedy. The design-patent number should be virtually or physically marked under § 287 to preserve pre-notice patent damages; the trade-dress claim should be flagged under § 1111 (by legend, since a shape rarely carries a symbol) to preserve profits and damages for trade-dress infringement. A common and expensive mistake is to assume that marking one of the two rights covers the other. It does not. A design-patent notice tells the world nothing about a trade-dress claim, and a trade-dress legend tells the world nothing about a patent. Each statute has its own notice requirement keyed to its own remedy, and the owner who wants both remedies must satisfy both.

A subtler interaction arises as rights expire. A design patent lasts fifteen years; trade dress can last forever. When the design patent expires, the patent marking should be updated (or removed from the virtual-marking page) — but the trade-dress legend should remain, because the trade-dress right survives the patent. Indeed, the moment of design-patent expiration is often precisely when the trade-dress claim becomes most important, because the patent monopoly that previously deterred copyists is gone and the owner is left to rely on its perpetual, distinctiveness-based right. A marking program that mechanically strips all IP notices when a patent lapses can inadvertently weaken the very right the company most needs going forward. (There is a doctrinal wrinkle here worth a footnote: a feature that was the central advance of a now-expired utility patent is, under TrafFix, strong evidence that the feature is functional and therefore not protectable as trade dress at all. The patent-to-trade-dress handoff works far more cleanly for ornamental design-patent features than for functional utility-patent features.) The lesson is that marking is not a one-time event but a living function that must track the independent life cycles of overlapping rights.

A Comparative Summary: Three Regimes, Three Logics

It is worth pausing to set the three regimes against one another, because their differences are exactly what a multi-right marking program must reconcile.

Patent (§ 287) Trademark / Trade Dress (§ 1111) Copyright (§§ 401–402)
Symbol / notice "Patent" / "Pat." + number, or virtual-marking URL ® (or statutory phrase); legend for trade dress © (or ℗ for sound recordings) + year + owner
Mandatory? No, but skipping it forfeits back-damages No, but skipping it can forfeit profits/damages No; copyright and remedies survive omission
What notice does Constructive notice unlocks pre-notice damages Constructive notice of registration unlocks profits + pre-notice damages Defeats innocent-infringer reduction, protecting statutory-damages award
Over-marking penalty Real: false marking under § 292 (intent to deceive) Quasi: deliberate misuse of ® = unclean hands Essentially none for good-faith ©
Who bears the burden Patentee (after infringer's burden of production) Mark owner Copyright owner
Key authority Dunlap, Arctic Cat, Amsted, Forest Group Wal-Mart, Two Pesos, TrafFix, § 1117 Berne; § 504(c)

Four observations follow from the table.

Is marking mandatory? None is strictly mandatory in the sense that failure forfeits the underlying right. But the practical pressure differs by an order of magnitude. Patent marking is the most coercive: skip it on a sold product and you lose back-damages outright. Trademark registration notice is moderately coercive: skip ® and you risk losing profits and pre-notice damages unless you can prove actual notice. Copyright notice is the least coercive: skip it and you keep your copyright and your remedies, losing only the ability to foreclose the innocent-infringer reduction.

What does notice unlock or protect? Patent marking unlocks pre-notice damages via constructive notice. Trademark ® unlocks (or protects) profits and pre-notice damages via constructive notice of registration. Copyright © protects the full statutory-damages award by defeating innocent infringement.

What is the penalty for over-marking? Here the regimes diverge most sharply. Patents have a genuine over-marking offense — false marking under § 292, with its intent-to-deceive element and (post-AIA) competitive-injury private right. Trademarks have a quasi-offense: deliberate misuse of ® on an unregistered mark, which can bar relief as unclean hands. Copyright has essentially no over-marking penalty; placing © on a work in which you in good faith claim copyright carries no analogous sanction. (Fraudulent registrations are a separate matter, governed by 17 U.S.C. § 411(b).)

Who bears the burden in litigation? Across the board, the rights holder generally bears the burden of proving notice compliance — that products were marked (patent), that ® was displayed or actual notice given (trademark), that © appeared on the accessed copies (copyright). This is why documentation is the connective tissue of every marking program.

Marking in Litigation: How Each Regime Plays Out in Court

The marking rules look like housekeeping until a lawsuit makes them dispositive. Seeing how each regime surfaces in litigation sharpens the case for getting marking right before any dispute arises.

In patent litigation, marking compliance is a damages-phase battleground. The accused infringer, eager to shrink its exposure, scrutinizes the patentee's marking history to argue that constructive notice never attached — pointing to unmarked units, unmarked licensee products, or a defective virtual-marking page — so that damages can run only from a much later actual-notice date, often the filing of the complaint. Under Arctic Cat, the infringer carries an initial burden of production to identify specific unmarked products it contends practice the patent; but once it does, the patentee must prove those products were in fact marked or that marking was excused. A patentee that cannot produce dated specimens, mold specifications, packaging records, and (for virtual marking) archived web pages is in a weak position. Because damages are usually a reasonable royalty or lost profits running from the start of infringement (see, e.g., LaserDynamics, Inc. v. Quanta Computer, Inc., 694 F.3d 51 (Fed. Cir. 2012)), a marking cut-off can lop years off the recovery. These dynamics fit within the broader damages analysis covered in our Comprehensive Guide to Patent Infringement Litigation.

In trademark and trade-dress litigation, the § 1111 question usually appears when the plaintiff seeks the defendant's profits and its own damages under § 1117. A defendant who can show the plaintiff never displayed registration notice will argue that monetary recovery is barred for the pre-notice period unless the plaintiff proves the defendant had actual knowledge of the registration. The plaintiff's display of ® (or a registered trade-dress legend), backed by dated specimens, forecloses the argument. Separately, defendants sometimes raise the plaintiff's misuse of ® as an unclean-hands defense — contending the plaintiff used the registration symbol on unregistered goods to deceive the marketplace; that defense bites only where the misuse was deliberate, but it underscores why scrupulous, registration-accurate use of ® matters. Liability itself turns on a separate likelihood-of-confusion analysis discussed in Navigating the Maze of Trademark Confusion, but the marking question controls how much money is on the table once liability is found.

In copyright litigation, the © notice surfaces when the defendant pleads innocent infringement to shrink a statutory-damages award under § 504(c). A plaintiff whose published copies bore a proper notice, and to which the defendant had access, defeats that plea under § 401(d)/§ 402(d), preserving the full $750–$30,000 (or up to $150,000 for willful infringement) per-work range. A plaintiff who omitted notice keeps its copyright and its remedies but hands the defendant a ready argument for the $200-per-work floor. Because timely registration is a separate prerequisite for statutory damages and attorneys' fees under § 412, the disciplined plaintiff pairs prompt registration with consistent notice — two cheap steps that together maximize leverage.

The common thread: in every regime, marking decisions made years earlier determine the size of the recovery years later, and in every regime the rights holder bears the burden of proving it did the right thing. That is the practical case for the program that follows.

Building a Unified Marking-Compliance Program

A company that owns several types of IP across many products cannot run three disconnected marking efforts. It needs one coordinated program with a recurring rhythm: inventory, decide, implement, document, monitor, correct.

Step one: inventory every right and every product

Start with a complete map. For each product, identify every IP right that touches it: utility patents, design patents, registered and unregistered trademarks and trade dress, and copyrighted elements (artwork, manuals, software, packaging text). Note the status and key dates of each — issuance dates, expiration dates, registration numbers, registration dates, and the goods/services covered. This inventory is the single source of truth that drives every marking decision, and it is the first thing litigation counsel will ask for.

Step two: decide the marking method for each right

For patents, choose physical or virtual marking. Virtual marking is usually the better choice for any product with more than a couple of patents, or a portfolio that changes over time, because it lets you update a web page instead of re-tooling a mold. Physical marking can be fine for a small, stable patent set on a long-cycle product. For trademarks, the decision is mostly automatic: ® on registered marks and registered trade dress (via symbol or legend), TM/SM on unregistered and pending marks. For copyright, the decision is also automatic: a proper © (or ℗ for sound recordings) on every published work.

Step three: implement consistently

Consistency is the recurring theme and the recurring failure point. Patent marking must reach "substantially all" units, continuously. Trademark ® should appear on at least the most prominent uses of the mark on goods and packaging (you need not festoon every appearance, but the lead uses should carry it). Copyright © should appear on each published work. The single best way to make this stick is to build the markings into the artifacts that generate the product — artwork files, mold specifications, packaging templates, and software build pipelines — so that compliance is the default rather than a manual step someone can forget under deadline.

Step four: document everything

In every regime the rights holder bears the burden of proving compliance, so contemporaneous documentation is decisive. Keep dated samples and photographs of marked products and packaging; archive the virtual-marking web page (with timestamps) every time it changes — a public web-archive snapshot is ideal; retain the marking covenants in license agreements and the records of licensee-compliance audits; keep specimens of ® and TM usage together with the registration certificates that justify them; and keep copies of published works showing the © notice in place. When a damages cut-off is litigated years later, this file is what carries the day. (For litigators, authenticating those archived web pages and product photos is its own discipline — see Capturing the Web -- A Practitioner's Guide to Authenticating Website Screenshots as Evidence in Federal Court.)

Step five: monitor and update

IP status changes constantly. Patents issue, expire, and get invalidated; trademarks register, renew, and lapse; new products and new copyrighted materials appear. Assign clear ownership of the program, set a recurring audit calendar (at least annual, more often for fast-moving lines), and use docketing or automated alerts to flag patent expirations and registration-renewal deadlines. The virtual-marking page in particular needs an owner who keeps it accurate and online, because a stale or unavailable page can both fail to give constructive notice and drift toward false marking — the worst of both worlds.

Step six: correct deficiencies promptly and in good faith

When an audit reveals a gap — unmarked patented units, ® on a mark whose registration lapsed, a missing copyright notice — fix it quickly and document the fix. For patents, begin marking and consider whether the actual-notice clock can be started. For trademarks, immediately stop using ® on anything no longer registered (switching to TM if common-law rights are claimed) and correct any registered-goods mismatch. For copyright, add the notice to future printings. Courts and the USPTO treat genuine, documented, good-faith correction far more sympathetically than ongoing neglect — and in the false-marking and improper-® contexts, the absence of deceptive intent is often outcome-determinative.

Worked Example: The "Acme Lumina" Smart Lamp

Abstractions become clear when applied to one product. Meet the Acme Lumina, a hypothetical premium smart desk lamp sold by the (invented) Acme Corp. The Lumina carries, all at once, the full spectrum of IP rights — making it an ideal vehicle for showing how a coordinated marking program looks in practice. Every fact below is illustrative.

Acme owns four kinds of protection on the Lumina:

  • A utility patent (U.S. Patent No. 11,234,567) covering the lamp's adaptive dimming circuit — a functional invention.
  • A design patent (U.S. Patent No. D987,654) covering the lamp's distinctive sculptural arm and base — an ornamental design.
  • Registered trade dress (a federal registration) in the overall silhouette of the lamp, which has acquired secondary meaning after several years of prominent sales and advertising.
  • Copyright in (a) the printed user manual and packaging artwork and (b) the software user interface displayed on the lamp's companion app.

Here is how each right is marked.

Utility and design patents. Because the lamp embodies two patents and Acme expects the portfolio to grow and the patents eventually to expire, Acme chooses virtual marking. The underside of the lamp base is molded with: "Patents: www.acme.com/patents." That page is freely accessible and lists, product by product: "Acme Lumina Smart Lamp — covered by U.S. Patent No. 11,234,567 and U.S. Patent No. D987,654." When patent 11,234,567 eventually expires, Acme updates the page rather than re-tooling the mold, relying on the § 292(c) safe harbor in the interim and on accurate web updates thereafter. This single step gives Acme constructive notice under § 287 for both the utility and the design patent, preserving its claim to pre-notice damages against any infringer — and it sidesteps the consistency trap, because the molded URL is on every unit that comes out of the mold. Acme's license to its overseas manufacturer contains an express marking covenant and audit right, so licensee units are marked too, closing the Maxwell gap.

Trademark and trade dress. The brand name "Lumina" is federally registered, so the lamp's nameplate and the packaging display "Lumina®." Because the lamp's silhouette is registered trade dress that cannot practically carry a symbol, Acme adds a packaging legend and a website statement: "The configuration of the Acme Lumina lamp is a registered trademark of Acme Corp." That display of registration notice satisfies 15 U.S.C. § 1111, preserving Acme's ability to recover the infringer's profits and its own damages from the start of any infringement rather than only from the date it can prove a copyist actually knew of the registration. Acme is careful to use ® only for the registered name and registered trade dress; a newer sub-brand still in the application stage is marked "FlexBeam™," not "FlexBeam®," to avoid improper-registration-symbol exposure and the unclean-hands defense it invites.

Copyright. The user manual and packaging artwork each carry "© 2026 Acme Corp." on the title page and the packaging footer. The companion app's "About" screen displays "© 2026 Acme Corp. All rights reserved." These notices are optional under the post-Berne regime, but they are valuable: if a competitor copies the manual or clones the app interface, Acme's visible © forecloses the copyist's innocent-infringer defense under § 401(d)/§ 402(d), protecting the full statutory-damages award — assuming Acme registered the works in time under § 412. The encircled-P symbol is not needed here because there is no sound recording; if the Lumina shipped with a copyrighted chime or audio guide fixed on a phonorecord, that would carry "℗ 2026 Acme Corp."

Documentation. Acme keeps a dated specimen of each marking: a photo of the molded patent URL, a timestamped archive of every version of the patents web page, a sample box showing "Lumina®" and the trade-dress legend, the trademark registration certificates, the copyright registration certificates, and copies of the manual and app screen bearing the © notice. When, two years later, Acme sues a copyist that has been selling knock-off lamps, this file lets Acme prove constructive notice for the patents, registration notice for the trademark and trade dress, and proper copyright notice — establishing the earliest possible damages start date for every right at once.

The example shows the payoff of coordination. One product, run through one program, has its patents virtually marked, its registered marks and trade dress flagged with ® and legends, its newer mark properly limited to TM, and its copyrighted works noticed with © — each notice doing the specific job that its statute rewards, and each backed by the documentation the rights holder will need to carry its burden in court.

International Marking Considerations

Most products are sold across borders, and marking that is sensible in the United States may be ineffective, unnecessary, or even unlawful elsewhere. A truly comprehensive program accounts for the major foreign variations without trying to satisfy every jurisdiction on a single label.

For patents, many countries do not impose a U.S.-style damages-marking requirement at all, so the constructive-notice logic of § 287 is largely a U.S. concern. But the false side of marking is taken seriously in many jurisdictions: marking a product as patented in a country where it is not patented, or after the local patent has expired, can constitute an offense or an unfair-competition violation. The United Kingdom, for example, treats false patent marking as actionable but recognizes a defense for diligent, good-faith use of a "patent — see [website]" style notice that directs the public to current information. This is one more reason virtual marking has become attractive for global sellers: a single web page can list, per product and per country, exactly which patents are in force where, allowing the product itself to bear only a neutral URL rather than country-specific claims that may be true in one market and false in another.

For trademarks, the ® symbol is internationally understood, but the rule that it may be used only where a valid registration exists is jurisdiction-specific. A mark registered in the United States but not in, say, Germany should not bear ® on goods sold into Germany based on the U.S. registration alone, because ® there represents a German (or EU) registration that does not exist; improper use can carry penalties under local unfair-competition law. The conservative practice for global products is to use ® only where the mark is registered in the destination market, to rely on TM elsewhere, or to manage the symbol through market-specific packaging. The TM and SM symbols, being informal and effectless, travel more safely — but they also carry less weight abroad than they do as evidence in U.S. secondary-meaning disputes.

For copyright, the Berne Convention's prohibition on formalities means notice is generally optional in member countries, just as it is in the United States. A © notice is widely recognized and harmless to include, and many sellers add the familiar "© [year] [owner]. All rights reserved." legend — the "all rights reserved" phrase being a vestige of the old Buenos Aires Convention that is now largely ceremonial but still commonly used out of habit and caution. Because © is cheap, recognized, and carries no downside, it is the easiest of the three to apply uniformly worldwide.

The practical synthesis for an international program: keep patent and trademark notices flexible and market-aware — favoring neutral virtual-marking URLs for patents and registration-accurate use of ® per market for trademarks — while applying copyright © uniformly. Coordinating this across markets is part of the larger brand-and-IP governance picture addressed in Brand Protection Online -- A Strategic Guide for Businesses.

Frequently Asked Questions

Is patent marking legally required? No. Marking is never required to keep the patent itself. But if you sell a physical product embodying the patent and do not mark it (physically or virtually), 35 U.S.C. § 287(a) bars you from recovering damages for infringement that occurred before you gave the infringer actual notice. Marking gives constructive notice and lets damages run from the start of the infringement.

My patent only covers a method — do I need to mark anything? Generally no. The marking requirement attaches to tangible articles that embody the patent. A patent that claims only a method or process, with no product to stamp, is not subject to § 287's damages limitation (American Medical Systems v. Medical Engineering Corp.). The same is true of a non-practicing entity that sells nothing embodying the patent. Be careful, though, where a patent has both apparatus and method claims and you sell an embodying product.

Can I put the ® symbol on my mark while my application is still pending? No. The ® symbol and the statutory phrases under 15 U.S.C. § 1111 may be used only once the federal registration certificate has actually issued, and only for the goods or services covered by the registration. That includes the window after a Notice of Allowance but before registration. Before registration, use ™ (for goods) or ℠ (for services). Deliberate misuse of ® can bar relief as unclean hands.

Do I still need a copyright notice if notice is optional? Yes, you should use one. For works published on or after March 1, 1989, omitting © does not forfeit your copyright, but a proper notice on the copies a defendant accessed defeats the "innocent infringement" defense under 17 U.S.C. §§ 401(d) and 402(d), protecting the full size of a statutory-damages award. It costs nothing and only helps.

What is the difference between ™ and ®? ® may be used only with a federally registered mark and gives statutory registration notice under § 1111, which conditions recovery of profits and damages. ™ (and ℠ for services) may be used with any claimed mark, registered or not; it has no statutory effect but signals a claim of rights and helps build the secondary-meaning record that unregistered trade dress in particular requires under Wal-Mart v. Samara Bros.

Can I get in trouble for marking too much? In patents, yes: marking with a patent that does not cover the product, or that has expired, with intent to deceive, violates 35 U.S.C. § 292 — though note the § 292(c) safe harbor for patents that previously covered the product and have since expired, and that after the America Invents Act only the United States may collect the per-article penalty (a private plaintiff needs a competitive injury). In trademarks, deliberate misuse of ® can bar relief as unclean hands. In copyright, good-faith use of © carries no analogous penalty. The cure across all three is to mark accurately and keep the record straight.

Whatever happened to all those false-marking lawsuits? They were a real phenomenon. After Forest Group v. Bon Tool held in 2009 that the § 292 penalty accrues per falsely marked article, and because anyone could sue and keep half the penalty, a wave of "marking troll" suits hit companies selling products marked with expired patents. The 2011 America Invents Act ended the boom by abolishing the qui tam bounty, limiting private suits to plaintiffs with a competitive injury, and adding the expired-patent safe harbor. The docket largely dried up within a year.

How does marking work for trade dress, which I cannot stamp with a symbol? Give notice through packaging legends, hangtags, and website statements rather than a symbol on the article. For registered trade dress, a legend reciting that the configuration is a registered trademark satisfies § 1111. For unregistered trade dress, consistently asserting trade-dress rights helps establish secondary meaning, which product-configuration trade dress always requires under Wal-Mart v. Samara Bros.

Key Takeaways

Marking is the public-notice machinery of intellectual property, and although it is rarely mandatory in the strict sense, getting it right is one of the highest-leverage, lowest-cost things an IP owner can do. Patent marking under § 287 unlocks pre-notice damages, must reach "substantially all" products and licensees consistently, and is policed on the other side by the false-marking statute, § 292 — a regime reshaped by Forest Group v. Bon Tool and then tamed by the America Invents Act. Trademark and trade-dress notice under § 1111 conditions the recovery of profits and damages; the ® symbol is reserved strictly for issued registrations, while TM/SM are free, effectless-but-strategic signals for unregistered and pending marks and useful evidence of the secondary meaning that configuration trade dress always demands. Copyright notice under §§ 401–402 is optional after Berne but protects the full statutory-damages award by defeating innocent infringement. A product carrying several of these rights needs one coordinated program — inventory, decide, implement, document, monitor, correct — and, above all, the documentation to carry the rights holder's burden of proving compliance when enforcement day finally arrives.

For the deep mechanics of patent marking specifically — actual-notice case law, virtual-marking page design, and licensee compliance — see Understanding Patent Marking Requirements -- A Practical Guide for Practitioners and Entrepreneurs.

Related Articles

This article is provided by mclaw.io for general informational purposes only. It is not legal advice, does not create an attorney-client relationship, and may not reflect the most current legal developments or the law of your jurisdiction. Marking decisions can have significant consequences for the recovery of damages and profits, and you should consult qualified intellectual property counsel before adopting or changing a marking program.