Cancelling a Registration at the TTAB: A Petitioner’s Checklist

By ·

A federal registration is formidable but never untouchable: under Section 14 of the Lanham Act it can be cancelled when it should never have issued or no longer deserves to stand. This checklist walks a challenger and its counsel through cancelling a registration at the Trademark Trial and Appeal Board, from choosing among the three venues—a TTAB petition, a Section 37 court counterclaim, or the USPTO's ex parte expungement and reexamination—to reading the registration's vital signs, matching grounds to timing, proving entitlement to a statutory cause of action after Lexmark, pleading through ESTTA, and prosecuting the case to a decision. A field guide to the grounds explains genericness, functionality, abandonment, the demanding post-In re Bose fraud standard, the new Section 1064(6) nonuse theory, likelihood of confusion, descriptiveness, deceptiveness, and dilution—each with WHY notes, trap warnings, and worked hypotheticals. It flags the cheaper Trademark Modernization Act routes for pure nonuse and the Section 18 power to restrict rather than cancel. Anticipated defenses, a costs-and-timeline map, and a common-mistakes list round out the playbook. Verify current TBMP guidance, 37 C.F.R. Part 2, and fees before filing.

Intellectual Property → Trademark Litigation | Published 28 June 2026 | rightsy.io

A federal trademark registration looks like a fortress. It carries a presumption of validity, a presumption of ownership, and a presumption of the exclusive right to use the mark nationwide. It can mature into something close to invulnerable through incontestability. And it sits on a public register that every examiner, every competitor, and every search tool treats as the official record of who owns what.

But a fortress built on bad ground can be condemned. Congress wrote Section 14 of the Lanham Act, 15 U.S.C. § 1064, precisely so that a registration that never should have issued—or that should no longer stand—can be removed from the register by a party who is harmed by its continued existence. The instrument for doing that at the U.S. Patent and Trademark Office is a petition to cancel, filed with the Trademark Trial and Appeal Board (TTAB or the Board).

This checklist takes you from the very first strategic question—is the Board even the right place to fight?—through reading a registration's vital signs, matching grounds to timing, proving you are entitled to bring the case at all, drafting and filing the petition, and prosecuting it to a final decision. It is built for two readers at once: the business owner trying to understand whether a blocking registration can be cleared out of the way, and the practitioner who has to actually file the thing. Use the checkboxes as a worklist; read the WHY notes and TRAP warnings to understand what the checkboxes are protecting you from.

One orientation point before the first box, because it governs everything that follows.

What the Board can—and cannot—do for you

The TTAB decides exactly one question: the right to register. It is an administrative tribunal, not a court of general jurisdiction. That single fact reshapes your whole strategy.

So if your real problem is that a competitor is selling knock-offs and pocketing your customers, cancelling its registration will not, by itself, make it stop. It only strips away the federal paperwork and the presumptions that come with it. The competitor may still own common-law rights in whatever it actually uses. (For how those unregistered rights work and where they run out, see the three tiers of trademark rights.)

The flip side is that a cancellation is often enough. If a stale or overbroad registration is the only thing blocking your own application, or the only basis for a cease-and-desist threat against you, vacating it can solve the whole problem—quietly, cheaply, and without a courtroom.

TRAP — the Board is not a scrimmage. It is tempting to treat a TTAB proceeding as low-stakes practice before the "real" court case. Do not. In B&B Hardware, Inc. v. Hargis Industries, Inc., 575 U.S. 138 (2015), the Supreme Court held that a Board determination on likelihood of confusion can have issue-preclusive effect in later district-court infringement litigation when the usages adjudicated are materially the same as those before the court. Lose a contested issue at the TTAB and you may have lost it for good. Plan the Board case as if a judge and jury were watching, because eventually one may be reading the opinion.


Phase 1 — Pick the battlefield: Board, court, or examiner

Before you draft anything, decide where the fight belongs. There are three distinct ways to attack a registration, and choosing the wrong one wastes money and time.

WHY / TRAPS. Section 1119 is not a freestanding grant of jurisdiction. You cannot walk into federal court with a complaint whose only purpose is "please cancel this registration"—there must be an independent case or controversy for the court to be hearing in the first place. (Courts have split on the precise contours, but the safe assumption is that § 37 modifies an existing action; it does not create one.) Conversely, do not default to a full Board trial for a problem the examiner can solve. If the registrant simply never used the mark on most of its listed goods, the TMA expungement route can clear those goods for a few hundred dollars without a single deposition.

Reconnaissance pays here. Before committing to a forum, learn who you are dealing with. Pull the registrant's full prosecution history, every assignment in the chain of title, and any oppositions or cancellations it has filed or defended. Rightsy's TTAB proceedings database lets you see whether your target is a serial filer with a settlement playbook or a one-registration small business, and Rightsy's assignment records reveal whether the mark has changed hands in ways that may themselves be a ground (more on assignment-in-gross and improper intent-to-use transfers below). Knowing the adversary's history shapes whether you petition, sue, or send the examiner after them.


Phase 2 — Read the registration's vital signs

Every cancellation strategy turns on two numbers and one status flag: how old is the registration, is it incontestable, and which grounds are still on the table? Get these wrong and you may plead a theory the statute closed off years ago.

WHY / TRAPS. The first analytical job in every cancellation is matching grounds to timing, and there is a subtle trap that the word "incontestability" tends to hide.

TRAP — the five-year clock runs from registration, not from the § 15 affidavit. Practitioners loosely say a registration is "incontestable" and therefore safe from descriptiveness challenges. But for cancellation purposes, what slams the door on the registrability grounds is the registration simply turning five years old under § 1064(1)—whether or not the owner ever filed a § 15 affidavit. A registrant who never perfected incontestability is still beyond a descriptiveness or surname attack once five years have passed, because those grounds are not listed in § 1064(3). Do not lull yourself into thinking "they never filed Section 15, so descriptiveness is still open." It usually is not. (The § 15 affidavit matters enormously for the conclusive-evidence defense in court under § 1115(b)—see what federal registration actually buys you—but it is not the gatekeeper for § 1064 grounds.)

TRAP — keep three nonuse-adjacent theories distinct. Nonuse (the mark was never used, or not in use as of a relevant date) is not the same as abandonment (use began and then stopped, with intent not to resume), which is not the same as fraud (a knowing, material lie to the USPTO). They have different elements, different burdens, and different remedies. Pleading them as if they were interchangeable is a classic rookie error.

A note on the Supplemental Register. If your target sits on the Supplemental Register rather than the Principal Register, the calculus shifts: there is no incontestability and no § 7(b) presumptions, descriptiveness is not a ground (a Supplemental Register mark is descriptive by definition), but genericness, abandonment, nonuse, fraud, functionality, and § 2(d) priority all still apply. Confirm which register you are attacking before you choose a theory.


Phase 3 — Prove you are entitled to bring the case

You can have the best genericness theory in the world and still be thrown out at the threshold if you cannot show you belong in the dispute. The Board calls this entitlement to a statutory cause of action. It used to be called "standing," and many practitioners and opinions still use that word, but the label changed for a reason.

WHY / TRAPS. Entitlement is construed liberally, but it is a genuine element of the case—not a formality. State it affirmatively in the petition and back it up.

TRAP — "competitor who wants to use the word" is a real interest for public-interest grounds. For genericness and functionality, the public has a stake in keeping common words and useful features free, so any competitor kept out of the market by the registration generally has entitlement. You do not need a confusingly similar mark of your own to attack a generic registration; you need to want to use the word the registration is hoarding.

TRAP — do not manufacture a stake you do not have. If the registration does not actually block you, the Board may find you have no real interest. Run the analysis honestly: a clearance search through Rightsy's trademark and logo search will show you, in black and white, whether the target registration sits in your path—and the resulting refusal or conflict is itself the evidence of your entitlement.


A field guide to the grounds

Most of the intellectual work in a cancellation is choosing and building the ground. Below is a working guide to the theories you are most likely to use, each with its elements, its proof, its traps, and a worked hypothetical. Treat these as modules: a strong petition often pleads two or three in the alternative.

A standing piece of strategic advice runs through all of them: lead with objective grounds. Genericness, functionality, abandonment, and nonuse can often be won on a paper record—documents, dictionary entries, the registrant's own filings, a clean survey. Fraud, which turns on a human being's hidden intent, almost never resolves on paper and usually has to wait for trial. Build the case you can win on summary judgment first.

Ground 1 — Genericness (the word became the thing)

A generic term names the category of goods, not their source, and no one may own the category. "Genericness" is a "becomes" ground: a mark valid at registration can later collapse into the common name through the public's own usage—the fate that overtook escalator, aspirin, cellophane, and thermos.

WHY / TRAPS. Genericness is the most public-spirited ground and one of the most durable—it is available at any time and cannot be cured. TRAP: a survey done badly is worse than no survey; design it to withstand a Daubert challenge before you spend the money. TRAP: distinguish genericness from mere descriptiveness; a descriptive mark can be saved by secondary meaning, but a generic term can never be rescued. (For the full ladder from generic to fanciful, see the Abercrombie spectrum.)

Worked example. ChillBrew, Inc. owns a registration for the word mark COLD BREW for coffee. A rival roaster, Harborview Coffee Co., is refused registration of "Harborview Cold Brew" because of it. Harborview pulls thirty competitor menus using "cold brew" as the ordinary name of the beverage, dictionary entries, and—best of all—ChillBrew's own packaging describing its product as "our signature cold brew." A Teflon survey shows 78% of coffee drinkers treat "cold brew" as a common name. Harborview has a textbook genericness petition and a clear real interest, because the registration is the only thing standing between it and its own application.

Ground 2 — Functionality (you cannot trademark a useful feature)

Trademark law refuses to let a brand owner monopolize a feature that competitors need in order to compete. Functionality is available at any time and, like genericness, cannot be cured.

WHY / TRAPS. Functionality is the doctrinal guardrail of trade-dress and product-configuration law. If you are attacking a registered product shape or color, this is frequently your best ground. For the broader doctrine—and the line between protectable look and unprotectable utility—see trade dress and the functionality bar. TRAP: once a feature is found functional under the TrafFix/Inwood test, the Board need not even reach competitive necessity; do not over-litigate alternative designs when the feature is essential on its face.

Worked example. GripRite, LLC registers, as trade dress, the spiral-fluted shape of its garden-hose nozzle. A competitor, Verdant Tools, discovers that GripRite's now-expired utility patent claimed the exact spiral flute as the mechanism that "improves grip and reduces slippage when wet." Under TrafFix, that expired patent is powerful evidence the shape is functional. Verdant petitions to cancel; the registration is highly vulnerable.

Ground 3 — Abandonment (use began, then stopped)

A registrant abandons a mark when use is discontinued with intent not to resume (§ 45, 15 U.S.C. § 1127). Abandonment is available at any time and is one of the most reliable objective grounds.

WHY / TRAPS. Abandonment is usually the stronger cousin of a "they barely used it" complaint (see Ground 5). TRAP: the three-year presumption is rebuttable; a registrant with documented, good-faith plans to relaunch can survive even a long gap. TRAP: "nonuse" of a registration that is being maintained on paper is suspicious—cross-check the registrant's § 8 and § 9 maintenance filings, because a use claim in a maintenance affidavit that contradicts your abandonment evidence can flip into a fraud theory. (For how those maintenance deadlines and affidavits work, see keeping your registration alive.) Use Rightsy's assignment records to trace whether the mark was sold off in a way that stripped its goodwill, and Rightsy's brand monitoring to catch the status changes—lapses, abandoned renewals, ownership shuffles—that signal a vulnerable registration.

Worked example. Meridian Motors registered ZEPHYR for automobiles, then discontinued the model line six years ago and now sells only parts under a different brand. A startup, Zephyr Mobility, wants the name for electric scooters and is blocked. Zephyr Mobility documents six years of no car sales—well past the three-year presumption—and Meridian's only response is a vague "we always meant to bring it back." That is the kind of uncorroborated intent the Board routinely rejects.

Ground 4 — Fraud on the USPTO (the hardest ground to win)

Fraud cancels a registration obtained or maintained through a knowing, material misrepresentation made with the intent to deceive the Office. It is the most dramatic ground and the one most likely to fail.

WHY / TRAPS. Bose itself is the cautionary tale: the registrant had renewed a registration claiming use on audio products it had stopped manufacturing (it was still repairing and shipping returned units), and the Federal Circuit found no fraud because there was no proof of intent to deceive. TRAP: do not bet the case on fraud. Treat it as a trial issue layered on top of a provable objective ground, not as your lead theory. TRAP: a mistaken or even careless overstatement of goods is not fraud; it is, at most, grounds to trim the registration.

Worked example. Lumen Labs swore, in its statement of use, that it was selling its AURORA mark on "cosmetics, candles, and home fragrance diffusers." Discovery in a cancellation reveals an internal memo: "We've never made diffusers—just list them so we're covered, and don't tell the examiner." That memo is the rare smoking gun showing both falsity and intent to deceive. Without it, the diffuser overclaim would merely get the goods struck; with it, the whole class is in jeopardy.

Ground 5 — Nonuse, including the new § 1064(6) theory

Nonuse means the mark was never validly used in commerce on the goods or services—distinct from abandonment, where use started and stopped.

WHY / TRAPS. TRAP — the use-in-commerce bar is famously low. A "they barely used it" theory is weak, because surprisingly little use qualifies: in Christian Faith Fellowship Church v. adidas AG, 841 F.3d 986 (Fed. Cir. 2016), the sale of two "ADD A ZERO" caps to an out-of-state buyer was held sufficient use in commerce. The lesson: if there was any genuine sale, drop "barely used" and pivot to genuine abandonment (use later stopped) or nonuse on the specific goods never sold. TRAP: § 1064(6) reaches goods never used; if use occurred and then ceased, that is abandonment, with its own three-year presumption.

Worked example. Summit Apparel registered TRAILHEAD listing thirty-five items—jackets, tents, water bottles, headlamps, and more—but has only ever sold T-shirts and caps. A competitor petitions under § 1064(6) to cancel the dead weight. The likely result is a restriction of the registration to the goods actually sold, freeing the rest of the list. (For why a bloated goods list is a liability in the first place, see goods, services, and the Nice classes.)

Ground 6 — Likelihood of confusion (within five years only)

If the registration conflicts with your earlier rights, you can cancel it on the same § 2(d) basis the examiner uses to refuse applications—but only inside the five-year window.

WHY / TRAPS. TRAP — B&B Hardware again. Because a Board confusion ruling can bind a later court, litigate the § 2(d) issue here with the same rigor you would bring to an infringement trial. A careless loss on confusion at the TTAB can haunt your district-court case.

Ground 7 — Descriptiveness and surname (within five years only)

WHY / TRAPS. TRAP: revisit the Phase 2 warning—these grounds close at the five-year mark from registration, not from a § 15 filing. If the registration is older than five years, do not plead descriptiveness; reach for genericness, abandonment, or nonuse instead.

Ground 8 — Deceptiveness, false connection, and § 2(a)/(b)/(c) (anytime)

WHY / TRAPS. These are underused but powerful against marks that trade on a false association. TRAP: do not confuse deceptiveness (§ 2(a), anytime, fatal) with deceptive misdescriptiveness (§ 2(e)(3), which can be cured by distinctiveness and is treated as a registrability-type ground).

Ground 9 — Dilution (within five years only)

Ground 10 — Misrepresentation of source (anytime, but narrow)


Phase 4 — Draft the petition and file through ESTTA

A petition to cancel is a pleading, not a brief. Its job is to state a plausible claim, not to prove the case—proof comes later, through testimony and notices of reliance. Keep it short, plain, and complete.

WHY / TRAPS. TRAP — do not over-plead the merits. Resist dumping your evidence into the petition; you do not get extra credit, and you may box yourself in. TRAP — watch the per-class math. Fees and challenges are reckoned per class; decide deliberately which classes are worth the fight. STRATEGY: lead with the objective grounds you can win on summary judgment and treat fraud as a trial-only issue. TRAP — the answer deadline matters to you, too: once the Board institutes the proceeding and the registrant fails to answer, you can move for default judgment—but calendar the registrant's deadline so you are ready to act the moment it passes.


Phase 5 — Run the proceeding (or the ex parte petition) to a decision

A TTAB cancellation is a real litigation with a fixed schedule and no live trial—the "trial" is a paper-and-deposition record. Here is the arc.

The ex parte alternative, in detail. If your only quarrel is nonuse, the Trademark Modernization Act gives you two examiner-driven petitions that skip the adversary process entirely:

WHY / TRAPS. TRAP — there is no live trial to save a thin record. If you did not get the admission or the document in discovery, it is not in the record at decision time. TRAP — confirm the current TMA windows. A temporary transitional rule that once allowed expungement against any registration over three years old has expired; the standard three-to-ten-year window now governs. STRATEGY — partial relief is still relief. Restricting a bloated registration to the goods actually used can clear your path just as effectively as a full cancellation, and it is far easier to obtain.


Anticipated defenses (plan for them before you file)

A well-prepared petitioner war-games the registrant's defenses in advance:

WHY / TRAPS. TRAP: before you file a § 2(d) cancellation, search the registrant's portfolio for a Morehouse-style prior registration that would neutralize your damage. Rightsy's TTAB and registration records make that portfolio review quick.


Costs, timeline, and the build-or-buy decision


Common mistakes (the petitioner's greatest-hits of self-inflicted wounds)


Primary authority

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

Related Resources

This checklist is general information, not legal advice. Whether a registration can be cancelled depends on its age, status, and the available grounds, and on facts specific to your situation. Consult qualified trademark counsel—or Rightsy's virtual trademark attorneys—before acting.

Read this article on Rightsy