Checklist · Intellectual Property

Cancelling a Trademark Registration at the TTAB: A Practical Checklist

Type
Checklist
Jurisdiction
National
Date
2025-04-13
Topic
Intellectual Property

A federal registration is a presumption, not a fortress, and this checklist walks a challenger and its counsel through dismantling one at the Trademark Trial and Appeal Board under Section 14 of the Lanham Act, 15 U.S.C. § 1064. It sequences the work: pick the right door among a Board petition, a Section 37 court counterclaim, and the Trademark Modernization Act's ex parte expungement and reexamination lanes; read the registration's age and incontestability status; map the available grounds against that timeline; establish entitlement to a statutory cause of action after Lexmark and Corcamore; and plead, file through ESTTA, and prosecute to judgment. It works through abandonment, genericness, functionality, and the punishing post-In re Bose fraud standard with invented hypotheticals, and flags the traps that sink petitions—betting on fraud, attacking the low use-in-commerce bar, and pleading constitutionally dead grounds. Verify current TBMP guidance, 37 C.F.R. Part 2, and USPTO fees before filing.

Cancelling a Trademark Registration at the TTAB: A Practical Checklist

Intellectual Property > Trademark Litigation | Published 13 April 2025 - Updated 3 August 2025 | UpLaw - uplaw.us

A federal trademark registration is a presumption, not a fortress. When a mark should never have issued—or should no longer stand—Section 14 of the Lanham Act lets a damaged party ask the Trademark Trial and Appeal Board to strike it from the register. This checklist takes you from choosing the right forum through prosecuting the proceeding, mapping the grounds against the registration's age, and separating the theories that win from the ones that only look like they should.

Keywords: petition to cancel, trademark cancellation, TTAB, Section 14, 15 U.S.C. § 1064, incontestability, abandonment, genericness, functionality, fraud on the USPTO, In re Bose, nonuse, TMA expungement, reexamination, entitlement to a statutory cause of action, Lexmark, ESTTA, primary significance, restriction remedy, B&B Hardware.

A registration certificate looks final. It arrives with a gold seal, a serial number, and a set of statutory presumptions that make the owner's life easier in every future fight. But the certificate is a rebuttable presumption of validity, not a decree of it. See 15 U.S.C. § 1057(b). The register fills up with marks that were never used, marks that describe rather than distinguish, marks that have gone generic in the mouths of ordinary consumers, and a smaller number of marks procured by lies to the examining attorney. Cancellation is the mechanism the Lanham Act supplies to clean house.

Under Section 14, 15 U.S.C. § 1064, "any person who believes that he is or will be damaged" by a registration may petition the Trademark Trial and Appeal Board (TTAB, or the Board) to cancel it. Two features of that grant define everything that follows. First, the Board adjudicates only the right to register. It does not decide the right to use a mark, it does not find infringement, and it awards no money. See 15 U.S.C. § 1067; B&B Hardware, Inc. v. Hargis Indus., Inc., 575 U.S. 138, 143–44 (2015). Second, the Board's jurisdiction is a creature of the Patent and Trademark Office; a cancellation is an administrative proceeding conducted entirely on paper, with no live hearing and no jury. Understanding those two boundaries is the difference between a petition that fits the forum and one that belongs somewhere else.

This checklist runs the proceeding in the order a careful practitioner actually works it: choose the door, read the clock, match the grounds to the clock, prove you are entitled to complain, plead it, file it, and prosecute it. Each phase carries a short list of action items and a longer discussion of why—the doctrine, the circuit texture, and the traps that turn confident petitions into dismissals. For cancellation raised as a counterclaim in federal court under Section 37, see Cancelling a Trademark in Federal Court: A Practical Guide. For the discovery machinery that supplies most of the trial record, see the Discovery Before the TTAB checklist and the comprehensive TTAB discovery guide. If you are challenging a published application rather than an issued registration, the companion path is opposition; see Opposing a Trademark Before the TTAB: A Practical Checklist.

Phase 1 — Choose the right door

  • [ ] Identify all three forums and pick deliberately: a TTAB cancellation petition under § 1064; cancellation as a counterclaim in district court under § 1119; or the USPTO's ex parte expungement or reexamination for pure nonuse under the Trademark Modernization Act.
  • [ ] For a clean "they never used it" complaint, price out the TMA lanes first—they are examiner-driven, far cheaper, and require no adversary.
  • [ ] If there is a live infringement or declaratory-judgment dispute, weigh a § 1119 counterclaim in court, which can also decide use, award damages, and grant an injunction.
  • [ ] If a court case and a Board case both exist, plan for coordination—stays, primary jurisdiction, and the preclusion consequences of letting one race ahead.

Why the door matters

Three doors lead to the same register, but they open onto very different rooms.

The TTAB petition is the workhorse. It is comparatively affordable, the Board's judges are trademark specialists, and the proceeding is confined to registrability. It is the correct forum when your only quarrel is with the registration and you do not need a use ruling, an injunction, or damages.

The Section 37 counterclaim, 15 U.S.C. § 1119, empowers a district court "in any action involving a registered mark" to order cancellation. The load-bearing phrase is "in any action involving a registered mark." Section 1119 is not a freestanding grant of jurisdiction. You cannot file a federal suit whose sole object is cancellation and expect it to survive; the statute rides along with an independent, justiciable controversy—typically an infringement claim, a declaratory-judgment action, or an unfair-competition dispute. When such a controversy already exists, folding cancellation into it can be efficient, because the same court can resolve use, confusion, damages, and the registration in one judgment. The venue comparison is developed at length in Cancelling a Trademark in Federal Court.

The TMA lanes—ex parte expungement under 15 U.S.C. § 1066a and ex parte reexamination under § 1066b, both live since December 18, 2021—are the newest and most underused option. Expungement targets a registration on the ground that the mark was never used in commerce on some or all of the listed goods or services; reexamination targets a use-based registration on the ground that the mark was not in use as of the relevant date (the filing date of the use claim, or the deadline to file it). Either petition is decided by an examining attorney, not litigated against an adversary, and the filing fee is a fraction of a full Board proceeding. For a challenger whose entire theory is "this registrant swept a hundred goods into one class and sold four of them," the TMA lane can accomplish in months, and for hundreds of dollars, what a cancellation would take two years and tens of thousands to attempt. These routes are the register-cleaning tools Congress built precisely because the deadwood problem had grown severe; the USPTO's own audit program (random proof-of-use demands on a slice of Section 8 and 71 maintenance filings) is the sister mechanism on the maintenance side. See Keeping a Trademark Alive: Post-Registration Maintenance, Renewal, and Incontestability.

Trap. Do not treat the choice as a formality, and do not treat the Board as a low-stakes warm-up for court. Under B&B Hardware, Inc. v. Hargis Industries, Inc., 575 U.S. 138 (2015), a Board determination of an issue—there, likelihood of confusion—can carry issue-preclusive effect in later infringement litigation when the ordinary elements of issue preclusion are met and the usages adjudicated are materially the same as those at stake in court. That cuts both ways. Win an issue at the TTAB and you may lock it in; lose one and you may be stuck with it. Choosing the Board is a strategic commitment, not a rehearsal.

Phase 2 — Read the clock: age and incontestability

  • [ ] Pull the registration and fix its age from the registration date, and its filing basis (use-based under § 1(a), intent-to-use under § 1(b), foreign-based under § 44, or Madrid under § 66(a)).
  • [ ] Determine whether the owner has secured incontestable status by filing a Section 15 affidavit after five years of continuous post-registration use. See 15 U.S.C. § 1065.
  • [ ] If the registration is within five years and contestable, the full menu of grounds is open.
  • [ ] If it is incontestable, confine yourself to the grounds that survive incontestability—the ones § 1064 preserves "at any time."
  • [ ] If a vulnerable young registration is approaching its fifth anniversary, move before the window closes on the registrability-based grounds.

Why the clock governs the grounds

Section 14 is structured as a timeline. For the first five years after registration, a petitioner may attack on essentially any ground that would have justified refusing the mark in the first place—the registration is fully "contestable." After five years, the statute slams several doors: 15 U.S.C. § 1064(1) and (3) permit cancellation "at any time" only on the enumerated fundamental grounds, and everything else drops away.

Incontestability compounds the effect. A registrant who files a Section 15 affidavit—declaring five consecutive years of continuous post-registration use, no adverse final decision, and no pending challenge—converts the registration's prima facie evidentiary status into conclusive evidence of the registrant's exclusive right to use the mark, subject to the defenses and defects Congress enumerated. See 15 U.S.C. §§ 1065, 1115(b). The Supreme Court settled the practical meaning of incontestability in Park 'N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189 (1985): an incontestable registration cannot be challenged on the ground that the mark is merely descriptive. That is the single most consequential thing incontestability does. It amputates the descriptiveness attack—and with it the surname and geographic-descriptiveness attacks—while leaving the fundamental grounds fully intact.

Note the two clocks are not identical. The five-year cutoff in § 1064 runs whether or not the owner ever files a Section 15 affidavit; the § 1115(b) conclusive-evidence effect requires the affidavit. In practice, a registration more than five years old is beyond descriptiveness-based cancellation regardless, and if the owner also filed under Section 15, it additionally enjoys the § 1115(b) shield in any use dispute. For how owners build and defend that status, see Keeping a Registration Alive: A Trademark Maintenance and Renewal Checklist and What Federal Trademark Registration Actually Buys You.

Hypothetical. Suppose Harborline registers for "software as a service featuring project-management tools" on the Principal Register. A competitor believes the mark is descriptive of a harbor-navigation product feature. If the registration is three years old, the competitor may petition to cancel on mere descriptiveness under § 2(e)(1). If the registration is six years old and Harborline filed a Section 15 affidavit at year five, that attack is dead on arrival—Park 'N Fly forecloses it. The competitor's only surviving routes are the anytime grounds: has Harborline abandoned the mark, is "Harborline" generic for the software, was the registration procured by fraud, is any claimed trade-dress feature functional? The clock did not change the facts; it changed which facts matter.

Phase 3 — Match the grounds to the clock

  • [ ] For a contestable registration, screen the full set: likelihood of confusion with your prior mark under § 2(d); mere descriptiveness without acquired distinctiveness under § 2(e)(1); primarily merely a surname under § 2(e)(4); geographic descriptiveness or misdescriptiveness under § 2(e)(2)–(3); deceptiveness and false suggestion of a connection under § 2(a); and dilution of a famous mark under § 43(c).
  • [ ] For an incontestable registration, confine yourself to the anytime grounds under § 1064(3): the mark is or has become generic; the registered matter is functional; the mark has been abandoned; the registration was obtained fraudulently or is being used to misrepresent the source of goods; plus the § 1064(6) nonuse ground for registrations at least three years old.
  • [ ] Do not plead a ground the Constitution has erased. The § 2(a) bars on disparaging and immoral/scandalous marks are dead letters after Matal v. Tam, 582 U.S. 218 (2017), and Iancu v. Brunetti, 588 U.S. 388 (2019).
  • [ ] Keep three easily-confused theories analytically separate: nonuse, abandonment, and fraud.

The contestable menu, and the constitutionally dead grounds

Within the five-year window, the strongest and most common attack is priority plus likelihood of confusion under Section 2(d): you owned confusingly similar rights first, the registrant's mark should never have registered over yours, and you are damaged by its continued registration. The multifactor analysis that governs is the same DuPont framework the examining attorney applies, drawn from In re E. I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973); for the full treatment, see Likelihood of Confusion in Trademark Law: A Brand Owner's Guide. Descriptiveness, surname, and geographic grounds round out the contestable menu—each available only while the window is open.

Two former grounds are simply gone. Section 2(a)'s prohibition on disparaging marks fell in Matal v. Tam, where the Supreme Court held that denying registration because a mark disparages offends the First Amendment's viewpoint-neutrality command. The immoral or scandalous clause fell two years later in Iancu v. Brunetti. Pleading either ground today is worse than futile; it signals that counsel has not read the last decade of Supreme Court trademark law. The saga of how these clauses rose and fell—and how the Washington football team's marks were first cancelled and then rescued—is chronicled in Disparaging Marks and the First Amendment: The Washington Redskins Saga After Matal v. Tam.

A cautionary coda: not every constitutional challenge to a Section 2 bar succeeds. The Federal Circuit held in In re Elster, 26 F.4th 1328 (Fed. Cir. 2022), that the § 2(c) names clause was unconstitutional as applied, but the Supreme Court reversed in Vidal v. Elster, 602 U.S. 286 (2024), holding that the content-based, viewpoint-neutral names bar is compatible with the First Amendment given trademark law's history. The lesson is to distinguish the two clauses the Court actually struck (disparagement and scandalousness) from the rest of Section 2, which remains good law.

Keep nonuse, abandonment, and fraud in separate boxes

The single most common conceptual error in cancellation practice is smearing three distinct theories into one accusation that "the registrant didn't really use this and lied about it." They are different claims with different elements and different burdens.

Nonuse asks whether the mark was ever used in commerce for the goods, or was in use as of a required date. It is a status fact keyed to a moment—application filing, statement-of-use filing, or, under § 1064(6), the three-year mark. It does not require intent of any kind.

Abandonment asks whether the owner discontinued use with intent not to resume. See 15 U.S.C. § 1127. It presupposes that use once existed and then stopped. Three consecutive years of nonuse is prima facie evidence, which shifts a burden of production—not persuasion—to the owner.

Fraud asks whether the owner knowingly made a false, material representation to the USPTO with intent to deceive. See In re Bose Corp., 580 F.3d 1240 (Fed. Cir. 2009). It is about a lie, proven to a punishing standard.

A single set of facts can implicate all three—but the pleadings, the evidence, and the odds are different for each. Collapsing them is how petitioners end up betting a winnable abandonment case on an unwinnable fraud theory. The register-clearing threats are catalogued from the owner's defensive perspective in The Life Cycle of a Trademark: How Brand Rights Arise, Strengthen, and Disappear.

Phase 4 — Confirm entitlement to a statutory cause of action

  • [ ] Establish entitlement to a statutory cause of action (the doctrine formerly called "standing"): a real interest in the proceeding and a reasonable belief in damage.
  • [ ] Tie that interest to something concrete—your own confusingly similar mark, prior use, a blocked application, or a genuine need to use a term the registration locks up (a descriptive or generic word your industry shares).
  • [ ] Plead entitlement in the petition; it is an element, not a formality, and a bare or conclusory allegation invites dismissal.

Why entitlement is a real element, liberally construed

For decades the Board spoke of "standing," and everyone understood it to mean a petitioner had to be more than an "intermeddler." The Federal Circuit modernized the vocabulary and the analytical frame in Corcamore, LLC v. SFM, LLC, 978 F.3d 1298 (Fed. Cir. 2020), holding that the statutory-cause-of-action framework the Supreme Court articulated for Lanham Act false-advertising claims in Lexmark International, Inc. v. Static Control Components, Inc., 572 U.S. 118 (2014), governs TTAB proceedings too. The petitioner must fall within the statute's zone of interests and show that its injury is proximately caused by the registration. The court took pains to say this is not a new, tougher test: it is "substantively the same" as the traditional "real interest plus reasonable belief of damage" formulation the Board had long applied. See Corcamore, 978 F.3d at 1305–06; Empresa Cubana Del Tabaco v. General Cigar Co., 753 F.3d 1270 (Fed. Cir. 2014).

In practice the bar is low but real. A commercial rival with a similar mark, a party whose application was refused or suspended because of the target registration, or a competitor who wants to use a descriptive or generic term the registration monopolizes all easily clear it. Notably, the petitioner need not own a mark or even have a competing product to attack a generic or functional registration—anyone with a genuine competitive interest in using the term or feature can be damaged by its improper monopolization. What the petitioner may not do is rest on a conclusory recital. Articulate the stake, connect it to the specific registration, and connect the registration to a plausible injury.

Hypothetical. Northwind Coffee, a regional roaster, wants to describe its darkest roast as "midnight" but discovers a competitor holds a registration for MIDNIGHT covering coffee. If Northwind petitions to cancel on genericness or mere descriptiveness, its entitlement is obvious: the registration blocks a word Northwind and every other roaster needs to describe their product. Northwind does not have to prove it was sued or even threatened. The blocked competitive use is the real interest, and the reasonable belief of damage follows from it.

Phase 5 — Build the substantive grounds

This is where cases are won or lost. Each ground has its own proof architecture. Below are the four anytime grounds most cancellation practice turns on, each with the governing standard, the evidence that moves the needle, and an invented illustration.

Abandonment: the strongest quiet weapon

  • [ ] Assemble evidence of three or more consecutive years of nonuse to trigger the statutory presumption.
  • [ ] Anticipate the owner's intent-to-resume rebuttal and gather facts that undercut it—no product, no marketing, no bona fide plans.
  • [ ] Distinguish genuine market use from token, residual, or sporadic use offered only to preserve rights.

Section 45 defines abandonment: a mark is abandoned when use is discontinued with intent not to resume, and "[n]onuse for 3 consecutive years shall be prima facie evidence of abandonment." 15 U.S.C. § 1127. The three-year presumption is the petitioner's friend—it shifts the burden of production to the registrant, who must then come forward with evidence of an intent to resume use within the reasonably foreseeable future. The burden of persuasion stays with the petitioner throughout. And courts have set the rebuttal bar modestly: the owner need only produce evidence from which a reasonable factfinder could conclude it intended to resume use. See ITC Ltd. v. Punchgini, Inc., 482 F.3d 135, 149–50 (2d Cir. 2007).

The battleground is usually whether "use" ever really stopped. Owners reach for scraps—an old invoice, a promotional t-shirt, a dormant website—to argue continuity. The case law is unkind to that maneuver when the scraps are contrivances rather than commerce. In Exxon Corp. v. Humble Exploration Co., 695 F.2d 96, 100–01 (5th Cir. 1983), the court found discontinued use where the owner's only activity was isolated shipments and select invoices generated solely to keep the mark alive. Sporadic, non-commercial licensing does not save a mark either. See Silverman v. CBS, Inc., 870 F.2d 40, 48–49 (2d Cir. 1989). And use on collateral goods after the core goods are gone typically will not do—promotional caps and shirts could not sustain rights in a fire-truck mark whose trucks had stopped selling. See Emergency One, Inc. v. Am. FireEagle, Ltd., 228 F.3d 531, 539 (4th Cir. 2000). The doctrine is developed further in The Life Cycle of a Trademark.

Hypothetical. Cascade Cycles registered its mark for bicycles in 2014, sold its last branded bike in early 2019, and shut its factory. In 2023 a startup wants the name and petitions to cancel for abandonment. The three-year gap (2019–2022) supplies the prima facie case. Cascade responds with a 2021 trade-show flyer and testimony that it "always meant to come back." Under the Punchgini standard that scintilla might survive summary judgment—but at trial the petitioner shows there was no product, no supplier, no financing, and no concrete relaunch plan, only a wish. A wish is not an intent to resume within the reasonably foreseeable future, and the registration falls. Contrast a registrant who paused production during a supply-chain disruption but kept tooling, retained distributors, and documented a dated relaunch schedule—there, the presumption is genuinely rebutted.

Why abandonment is the "quiet weapon": it is objective, it is often provable from the registrant's own documents and admissions in discovery, and it does not require proving anyone's dishonesty. It frequently succeeds where a flashier fraud theory built on the same facts would collapse.

Genericness: the primary-significance test

  • [ ] Define the genus of goods or services at issue, then ask what the relevant public primarily understands the term to name.
  • [ ] Gather dictionary entries, media and trade usage, competitors' use, and the registrant's own generic usage.
  • [ ] Consider a properly designed Teflon or Thermos survey of primary significance.

A generic term names the category, not the source; it can never function as a trademark, and a registration for a generic term is void at any time. The Lanham Act itself codifies the test: "The primary significance of the registered mark to the relevant public . . . shall be the test for determining whether the registered mark has become the generic name of goods or services." 15 U.S.C. § 1064(3). The Federal Circuit's two-step framework asks (1) what is the genus of goods or services, and (2) does the relevant public understand the term primarily to refer to that genus? See H. Marvin Ginn Corp. v. Int'l Ass'n of Fire Chiefs, Inc., 782 F.2d 987 (Fed. Cir. 1986). Courts often distill it to a slogan: a trademark answers "who are you?"; a generic term answers "what are you?" See Yellow Cab Co. of Sacramento v. Yellow Cab of Elk Grove, Inc., 419 F.3d 925, 929 (9th Cir. 2005).

The evidentiary sources are well worn. Dictionaries and encyclopedias show ordinary meaning. Trade and news media show how the industry and public talk. Competitors' generic use shows the term is a common name the market needs. Most damning is the registrant's own generic use—brochures, packaging, or SEC filings that treat the "mark" as a product category. Verb use is suggestive but not dispositive; the Ninth Circuit rejected the argument that GOOGLE is generic merely because people say "google it," insisting the inquiry is whether the primary significance to consumers is the category or the source. See Elliott v. Google, Inc., 860 F.3d 1151 (9th Cir. 2017). And a term generic for even one good or service in a listing is unregistrable as to that good. The Supreme Court refined the doctrine for the internet era in USPTO v. Booking.com B.V., 591 U.S. 549 (2020), holding that a "generic.com" term is not automatically generic if consumers perceive the composite as a source-identifier—an evidentiary question that often turns on survey proof.

The two survey formats, each named for the case that blessed it, are the Thermos survey (which asks how consumers would ask for the product) and the Teflon survey (a mini-lesson in brand-versus-common-name followed by a sorting exercise). Both trace to landmark genericide cases—King-Seeley's loss of THERMOS, King-Seeley Thermos Co. v. Aladdin Indus., Inc., 321 F.2d 577 (2d Cir. 1963), and, further back, Learned Hand's holding that ASPIRIN had become the public's name for the drug in Bayer Co. v. United Drug Co., 272 F. 505 (S.D.N.Y. 1921). For survey design and how to attack a flawed one, see Reading and Building Trademark Confusion Surveys: A Practical Guide to Expert Methodology, the confusion-survey design checklist, and Daubert and the Consumer Survey.

Hypothetical. A company registers CLOUDVAULT for online data-storage services and then, over five years, is quoted in the trade press calling any encrypted storage locker "a cloudvault," ships documentation describing "your cloudvault and other cloudvaults," and watches three competitors adopt the term. A petitioner defines the genus—encrypted cloud storage—and runs a Teflon survey in which a majority of IT buyers sort "cloudvault" as a common name rather than a brand. The registrant's own generic usage, plus the survey, plus competitor adoption, is a textbook primary-significance case. Genericness, like abandonment, wins on objective public perception, not on anyone's state of mind.

Functionality: the doctrine that trumps everything

  • [ ] Identify the specific feature claimed as a mark (product configuration, packaging, color) and ask whether it is essential to use or purpose or affects cost or quality.
  • [ ] Hunt for an expired or existing utility patent claiming the feature—strong evidence of functionality.
  • [ ] Mine the registrant's advertising for touts of the feature's utilitarian advantages.
  • [ ] Distinguish de facto functionality (every feature does something) from de jure functionality (the feature is there because it works better).

Functionality is trademark law's hard stop. A functional feature cannot be a trademark no matter how distinctive or famous it has become, because trademark law may not be used to monopolize useful product features—that is the domain of the patent system, with its limited term. The foundational test comes from Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 850 n.10 (1982): a feature is functional if it is essential to the use or purpose of the article or affects its cost or quality. The Court sharpened the doctrine in Qualitex Co. v. Jacobson Products Co., 514 U.S. 159 (1995) (color can be a mark only if non-functional), and delivered the modern capstone in TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 (2001), which held that an expired utility patent is strong evidence that the claimed features are functional, and that once utilitarian functionality is established there is no need to inquire into competitive necessity.

The Federal Circuit and the TTAB draw a semantic line worth keeping straight. De facto functionality—the trivial truth that a bottle holds liquid—does not defeat a mark. De jure functionality—the design is the way it is because it works better—does. See TrafFix, 532 U.S. at 33–34. The classic evidentiary checklist (utility patents, the owner's own touting of utilitarian advantages, alternative designs, and manufacturing economies) is the practitioner's toolkit. Product-configuration trade dress carries an additional handicap: under Wal-Mart Stores, Inc. v. Samara Bros., 529 U.S. 205 (2000), product design is never inherently distinctive and always requires secondary meaning—so a functionality attack and a distinctiveness attack often travel together. The full doctrine, from Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763 (1992) forward, is developed in Trade Dress Protection: Where Brand Identity Meets the Functionality Bar and What Can Be a Trademark: The Subject Matter of Trademark Law.

Hypothetical. GripPro registers the hexagonal cross-section of its tool handle as trade dress. A competitor petitions to cancel, produces GripPro's own expired utility patent claiming that the hex shape "prevents rolling and improves torque," and pulls GripPro's ads bragging that the shape "won't slip." Under TrafFix the expired patent is powerful evidence of functionality, and the advertising seals it: the shape is there because it works better, not to identify GripPro. The registration is cancelled on functionality without any need to run a survey or weigh alternative designs.

Fraud on the USPTO: the ground everyone overrates

  • [ ] Secure genuine evidence of a knowing, material misrepresentation made with intent to deceive—not merely an inaccurate statement.
  • [ ] Plead fraud with Rule 9(b) particularity: the who, what, when, where, and the specific false statement.
  • [ ] Recognize that overclaiming goods usually yields restriction, not cancellation, and that intent is rarely resolvable on a paper record.

Fraud is the ground clients ask for by name and counsel should approach with dread. The Federal Circuit reset the standard in In re Bose Corp., 580 F.3d 1240 (Fed. Cir. 2009), holding that fraud requires a knowing, false, material representation made with the intent to deceive the PTO, proven by clear and convincing evidence. The court's memorable phrasing is that fraud must be "proven to the hilt" and may not rest on speculation. Bose, 580 F.3d at 1243, 1245. Crucially, Bose overruled the lax standard the Board had used for the prior six years in Medinol Ltd. v. Neuro Vasx, Inc., 67 U.S.P.Q.2d 1205 (T.T.A.B. 2003), under which a "knew or should have known" negligence-flavored showing sufficed. After Bose, negligence is not fraud, gross negligence is not fraud, and even a materially false statement is not fraud absent proof of subjective deceptive intent. The elements are stated consistently across the circuits. See Covertech Fabricating, Inc. v. TVM Bldg. Prods., Inc., 855 F.3d 163, 174–75 (3d Cir. 2017); MPC Franchise, LLC v. Tarntino, 826 F.3d 653, 658 (2d Cir. 2016).

Because it sounds in fraud, the claim must be pleaded with particularity under Federal Rule of Civil Procedure 9(b): the specific representation, its falsity, its materiality, and facts supporting a strong inference of deceptive intent. See San Juan Prods., Inc. v. San Juan Pools, Inc., 849 F.2d 468, 472 (10th Cir. 1988). A petition that alleges fraud "on information and belief" without pleaded facts of intent will be stricken. And even when a registrant did overstate its goods—claiming use on twelve items when it used the mark on nine—the ordinary remedy is restriction of the registration to the goods actually used, not wholesale cancellation, unless the petitioner can prove the overclaim was a deliberate lie. That restriction remedy is the Bose-era safety valve, discussed in Phase 8. The library's dedicated treatment is Fraud on the USPTO, Nonuse, and Lack of Bona Fide Intent Claims (Practical Law).

Hypothetical. Lumen Labs files a use-based application swearing use of its mark on "skincare serums, cleansers, and toners," but at filing it had sold only serums. A petitioner cries fraud. Under Bose, the misstatement about cleansers and toners is not enough; the petitioner must prove Lumen knew the statement was false and intended to deceive the examiner—perhaps an internal email reading "we haven't launched cleansers yet, but list them anyway so nobody else can." Absent that smoking gun, the likely outcome is restriction of the registration to serums, not death. The strategic takeaway repeats across cancellation practice: do not bet the case on fraud. Plead the objective ground (here, nonuse or partial abandonment) as the lead, and treat fraud as a trial issue you may not reach.

The nonuse and bona-fide-intent grounds—and the trap beneath them

  • [ ] For intent-to-use registrations, investigate the absence of a bona fide intent to use the mark at filing—objective, documented intent, not a bare say-so.
  • [ ] Screen for improper pre-use assignment of an ITU application (assignment-in-gross of a mark not yet used).
  • [ ] Before running any "they barely used it" theory, remember how low the use-in-commerce bar sits.

An intent-to-use application filed under Section 1(b) requires a bona fide intention to use the mark in commerce, and the Federal Circuit has made clear that "bona fide" means objective evidence, not merely the applicant's after-the-fact assertion. See M.Z. Berger & Co. v. Swatch AG, 787 F.3d 1368, 1375–76 (Fed. Cir. 2015); Kelly Services, Inc. v. Creative Harbor, LLC, 846 F.3d 857, 863–64 (6th Cir. 2017). A registrant who filed dozens of ITU applications with no business capacity to use them, no product development, and no documentary trail of intent is vulnerable. The ITU system's mechanics—constructive-use priority, the notice of allowance, and the statement of use—are covered in Intent-to-Use Trademark Filings: A Practical Guide, and the assignment traps in Recording a Trademark Assignment with the USPTO: A Practical Checklist.

Now the trap. The § 1064(6) nonuse ground and abandonment both invite a tempting argument: "the registrant hardly used this mark, so it's not really used." Resist it. The threshold for use in commerce is famously low. In Christian Faith Fellowship Church v. adidas AG, 841 F.3d 986 (Fed. Cir. 2016), the sale of two "ADD A ZERO" hats to an out-of-state buyer was enough to establish use in commerce sufficient to support registration. If a couple of hats crossing a state line qualifies, "they only sold a little" is not a theory—it is a losing quibble. The winning attacks are no use (nonuse/expungement) or stopped use (abandonment), not small use.

Phase 6 — The TMA fast lanes for pure nonuse

  • [ ] For a mark never used on some or all goods, file an expungement petition under 15 U.S.C. § 1066a (available for registrations generally between three and ten years old, subject to the current transitional rules).
  • [ ] For a mark not in use as of the relevant date on a use-based registration, file a reexamination petition under § 1066b (available in the registration's first five years).
  • [ ] Support either petition with a reasonable investigation and documentary evidence of nonuse.
  • [ ] Consider the § 1064(6) cancellation nonuse ground as the adversarial alternative for registrations at least three years old.

The Trademark Modernization Act of 2020 built three new tools into the register-cleaning arsenal, all aimed at deadwood—registrations covering goods the owner never actually sold. Two are ex parte and cheap; one is a new inter partes ground.

Expungement, 15 U.S.C. § 1066a, lets any party (or the Director) ask the USPTO to remove some or all goods from a registration because the mark was never used in commerce on those goods. Reexamination, 15 U.S.C. § 1066b, addresses use-based registrations where the mark was not in use in commerce on or before the relevant date (the actual use date claimed, or the deadline for filing the use claim). Both are decided by an examining attorney after the petitioner submits a verified statement describing a reasonable investigation and the supporting evidence; there is no adversary, no discovery, and no trial. The § 1064(6) cancellation ground is the litigated cousin—it lets a party petition the Board to cancel a registration at least three years old on the basis that the mark has never been used in commerce on some or all goods.

For a challenger whose entire complaint is nonuse, these lanes are transformative. Expungement can strip a padded goods list for a few hundred dollars per class and a well-documented investigation, without ever engaging the registrant in litigation. A full Board cancellation, by contrast, means pleadings, disclosures, discovery, testimony periods, and briefing over roughly two years. Choose the ex parte lane when nonuse is the whole case; reserve the cancellation petition for when nonuse travels with other grounds (confusion, genericness, functionality) that the ex parte process cannot reach. The USPTO built these procedures precisely because the register had grown clogged with unused marks—the same concern that drives its post-registration audit program. See Keeping a Trademark Alive: Post-Registration Maintenance, Renewal, and Incontestability.

Hypothetical. A foreign apparel house holds a five-year-old U.S. registration listing "clothing, namely, shirts, pants, jackets, hats, socks, gloves, scarves, belts, and footwear," but a market investigation shows it has only ever sold shirts and jackets in the United States. Rather than file a full cancellation, the challenger files an expungement petition documenting the investigation—retailer searches, the brand's own U.S. e-commerce catalog, customs data—and asks the USPTO to strike the seven unused categories. The examiner issues an office action to the registrant, which cannot show use, and the registration is trimmed to shirts and jackets. Fast, cheap, and adversary-free.

Phase 7 — Plead the petition

  • [ ] Draft a petition to cancel satisfying Federal Rule of Civil Procedure 8(a): a short, plain statement of entitlement and of each ground for relief.
  • [ ] Identify the target registration by number, and specify the goods, services, and classes you challenge.
  • [ ] Plead each ground as a separate count, with the factual allegations that support it; plead any fraud count with Rule 9(b) particularity.
  • [ ] Keep the petition a pleading, not a brief—merits evidence comes later.

The TTAB applies the Federal Rules of Civil Procedure, adapted by the Trademark Rules in 37 C.F.R. Part 2 and the Trademark Trial and Appeal Board Manual of Procedure (TBMP). A cancellation petition must allege the petitioner's entitlement to a statutory cause of action and set out each ground in a short, plain statement that gives the registrant fair notice. See 37 C.F.R. §§ 2.111–2.112; Fed. R. Civ. P. 8(a). Notice pleading governs most grounds; fraud is the exception and demands particularity under Rule 9(b).

Structure the petition so each theory stands on its own count—Count I abandonment, Count II genericness, Count III functionality, and so on—so that the loss of one does not imperil the others and so the Board (and any reviewing court) can see the elements matched to allegations. Lead with the objective grounds that can be won on a paper record: genericness, functionality, abandonment, and nonuse are provable from documents, admissions, and surveys, and are candidates for summary judgment. Position fraud, if pleaded at all, as a trailing count you expect to try, because deceptive intent almost never resolves on the papers. And resist the urge to argue the merits in the petition; the pleading's job is to state claims, not to prove them. The Board's pleading practice is mapped stage by stage in the TTAB Practice Toolkit.

Phase 8 — File through ESTTA and pay the fee

  • [ ] File the petition electronically through ESTTA (the Electronic System for Trademark Trials and Appeals) at uspto.gov.
  • [ ] Pay the per-class government fee—confirm the current amount, which the USPTO periodically adjusts.
  • [ ] For a TMA route, file the expungement or reexamination petition with the verified statement, the reasonable-investigation description, and the nonuse evidence.
  • [ ] Docket the resulting institution order immediately.

Filing is mechanical but unforgiving on fees and deadlines. ESTTA is the mandatory portal; paper filing is permitted only in narrow, excused circumstances. The fee is assessed per class challenged, so a petition attacking a three-class registration costs three times the single-class fee—budget accordingly, and confirm the number on the current USPTO fee schedule rather than relying on any figure quoted in secondary sources, including this one. Once the petition is accepted, the Board issues an institution order that sets the entire schedule. Calendar it the day it arrives.

Phase 9 — Prosecute the proceeding to judgment

  • [ ] Calendar the institution order dates: the discovery conference, discovery period, disclosure deadlines, and the sequential testimony (trial) periods.
  • [ ] Hold the mandatory discovery conference and serve initial disclosures before seeking discovery.
  • [ ] Build the record under the Discovery Before the TTAB checklist, mapping each ground to specific interrogatories, document requests, and admissions.
  • [ ] Introduce evidence at trial by testimony declaration or deposition and by notice of reliance.
  • [ ] Move for summary judgment on any objective ground where the material facts are undisputed.
  • [ ] Track the docket on TTABVUE, and confirm the registrant files an answer—default risks judgment.

There is no live trial at the TTAB and no witness ever takes a stand before the judges. The "trial" is a set of sequential testimony periods during which each party submits evidence in writing: testimony by declaration or deposition transcript, and documentary evidence by notice of reliance (the vehicle for printed publications, official records, discovery responses, and the like). See 37 C.F.R. §§ 2.122, 2.123. That structure has a strategic consequence—the discovery record is the trial record. Admissions extracted in discovery, the registrant's own documents, and interrogatory answers frequently supply the decisive proof. A petitioner who treats discovery as preliminary skirmishing rather than trial preparation has misread the forum. Map every ground to specific requests: for abandonment, sales and shipping records year by year; for genericness, marketing materials and internal category references; for functionality, engineering documents and the advertising file; for nonuse, the specimen provenance and the first-use chronology. The discovery caps, the disclosure regime, and the mechanics are detailed in Discovery Before the TTAB: Interrogatories, Document Requests, and Admissions.

Two procedural realities deserve emphasis. First, summary judgment is a real path to victory on objective grounds. When the registrant's own documents establish three years of nonuse, or an expired utility patent establishes functionality, the Board can and does rule without a trial period. Second, watch the answer. If the registrant defaults, the petitioner can move for judgment—but the Board is protective of the register and generally requires the petitioner to establish its grounds even on default for many claims, so do not assume a defaulted registrant hands you an automatic win on every theory.

The restriction remedy: partial relief the Board can grant

Cancellation is not binary. Under Section 18, 15 U.S.C. § 1068, the Board may restrict or modify a registration rather than cancel it outright—narrowing the goods and services, adding limitations, or otherwise reshaping the registration to reflect the marketplace reality. This is the natural remedy when a registrant used its mark on some listed goods but not others, and it dovetails with the Bose overclaiming principle: a padded goods list ordinarily earns a trim, not a death sentence. A petitioner who wants everything should still plead in the alternative for restriction, and a petitioner whose evidence proves nonuse on only part of the listing should frame the requested relief precisely. The district-court analog under Section 37, 15 U.S.C. § 1119, gives courts the same modify-or-cancel discretion; see Cancelling a Trademark in Federal Court: A Practical Guide.

Review of the Board's decision

A losing party is not stuck with the Board's judgment. Section 21 of the Lanham Act, 15 U.S.C. § 1071, offers two review paths: a direct appeal to the U.S. Court of Appeals for the Federal Circuit on the administrative record, or a civil action in a federal district court under § 1071(b), where the parties may introduce new evidence and the court decides the registration question de novo on the enlarged record. The choice between them is consequential—the Federal Circuit reviews the Board's factual findings deferentially for substantial evidence, whereas a § 1071(b) district-court action lets a party cure an evidentiary gap that sank it at the Board. This is also where the B&B Hardware preclusion warning comes home to roost: the issues you litigate and lose here can follow you into court, so treat the Board proceeding as the main event it is.

Common mistakes that sink cancellation petitions

  • Filing at the wrong door. Choosing a Board petition when a Section 37 court counterclaim (with damages and an injunction) would serve better—or filing a full cancellation when a cheap TMA expungement would have trimmed the padded goods list.
  • Pleading registrability grounds against an incontestable registration. Mere descriptiveness, surname, and geographic-descriptiveness attacks are foreclosed after five years and after Park 'N Fly.
  • Betting the case on fraud. After In re Bose, negligence and even gross negligence are not fraud; deceptive intent must be proven to the hilt, and overclaiming usually earns restriction, not cancellation.
  • Running a "barely used it" theory. The use-in-commerce bar is low (Christian Faith Fellowship—two hats sufficed); attack no use or discontinued use, not small use.
  • Pleading constitutionally dead grounds. Disparagement (Matal v. Tam) and immoral/scandalous marks (Iancu v. Brunetti) are gone; do not resurrect them.
  • Neglecting entitlement. Entitlement to a statutory cause of action is liberally construed but is a real element; a conclusory allegation invites early dismissal under Corcamore and Lexmark.
  • Treating discovery as a warm-up. In a paper trial, the discovery record is the evidence; admissions and documents win the case.
  • Ignoring the preclusion stakes. Under B&B Hardware, a Board loss can bind you in later litigation—there is no such thing as a low-stakes issue.

Primary authority

Verify current TBMP guidance, 37 C.F.R. Part 2, the TMA rules, and USPTO fees before filing.

Related UpLaw Resources

Related Checklists

Related Guides

Related Articles

Related Toolkits

This checklist is general legal information from UpLaw (uplaw.us), not legal advice, and does not create an attorney-client relationship. Whether a registration can be cancelled turns on its age, status, filing basis, and the available grounds; consult qualified trademark counsel and verify current statutes, rules, TBMP guidance, and USPTO fees before acting.