Maybe you just received a cease & desist letter. Maybe the USPTO refused your application under Section 2(d). Maybe Amazon pulled your listing and you're trying to figure out why.
All three roads lead back to the same legal concept: likelihood of confusion. It's the single most important test in U.S. trademark law, and most founders, sellers, and designers don't fully understand how it works — until it's used against them.
This guide explains the test in plain English: where it comes from, what the 13 DuPont factors actually mean, why USPTO approval doesn't guarantee you're safe, and how to self-assess your own brand before someone else does it for you.
What "Likelihood of Confusion" Actually Means #
The Lanham Act — the federal statute that governs U.S. trademarks — uses the phrase as the bar for both refusing a trademark registration (Section 2(d), 15 U.S.C. §1052(d)) and finding trademark infringement (Section 43(a), 15 U.S.C. §1125(a)).
The statutory language asks one question: would a reasonable consumer, encountering both marks in the marketplace, be likely to be confused as to the source, sponsorship, or affiliation of the goods or services?
Three things about that question matter a lot:
- It's about likelihood, not certainty. The plaintiff doesn't need to prove anyone was actually confused — just that confusion is plausible.
- It's measured from the perspective of the "average consumer" — typically a hurried, distracted shopper, not a careful expert.
- "Confusion" doesn't just mean thinking the two products are literally the same. It includes thinking they're related, sponsored by the same brand, or affiliated.
That third point is why a clothing brand named "DISNEY APPAREL" can be blocked even though Disney itself doesn't sell apparel under that exact mark. Consumers might reasonably think Disney sponsored or licensed it. Confusion as to sponsorship is enough.
Where the Test Comes From: The DuPont Case #
The current framework comes from a 1973 decision of the U.S. Court of Customs and Patent Appeals (predecessor to the Federal Circuit): In re E. I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973).
That case set out 13 factors that decision-makers should consider when relevant. Critically, the court emphasized there is no single litmus test — every case weighs whichever factors are probative on its facts. Some cases turn on one factor; most turn on two or three.
The factors apply both at the USPTO (when an examining attorney decides whether to refuse a mark under Section 2(d)) and in federal court (when a judge decides whether to find infringement). The weight given to each can differ between the two forums.
The 13 DuPont Factors Explained #
Below is each factor in plain English, with the practical question it asks.
1. Similarity of the marks themselves #
How similar do the marks look, sound, and connote? Comparing the words/designs in sight, sound, meaning, and commercial impression. This factor is doing the heavy lifting in most cases.
2. Similarity of the goods or services #
Do the goods/services overlap, or are they "related" enough that consumers might think they come from the same source? T-shirts and hats are related. Software and SaaS services are related. Cars and dog food are not.
3. Similarity of established trade channels #
Are both marks sold through the same channels? Big-box retail vs. specialty boutiques vs. direct-to-consumer online — different channels reduce confusion risk.
4. Conditions of purchase (impulse vs. careful) #
Is the product an impulse buy ($3 lip balm) or a considered purchase (a $40K car)? Confusion is more likely on impulse buys because the consumer makes a quick decision.
5. Fame of the prior mark #
The stronger and more famous the senior mark, the wider its protection. Famous marks (Disney, Apple, Coca-Cola) get protection extending well beyond their registered classes.
6. Number and nature of similar marks already in use #
Is the senior mark surrounded by similar marks for similar goods? A "crowded field" weakens the senior mark's distinctiveness and lowers confusion risk.
7. Actual confusion (evidence of) #
Has any consumer ever actually been confused? Real-world evidence of misdirected purchases, customer service calls, or social media mix-ups is powerful. Absence of evidence after long coexistence cuts the other way.
8. Length of concurrent use without confusion #
If both marks have been used side-by-side for years without incident, that's evidence confusion isn't likely. Relevant in close cases.
9. Variety of goods on which the mark is used #
If the senior mark is used across many product categories, consumers may expect further expansion. Wider use = wider protection.
10. Market interface between applicant and prior user #
Has there been any consent agreement, license, or coexistence arrangement between the parties? Existing agreements can rebut or confirm confusion risk.
11. Right to exclude others #
How vigorously has the senior owner enforced its mark against third parties? An owner who has consistently policed the mark has stronger rights than one who has tolerated infringers.
12. Extent of potential confusion (de minimis vs. substantial) #
How many consumers, in absolute terms, are likely to be confused? Niche marks with tiny audiences may produce only de minimis confusion.
13. Any other established fact #
A catch-all. Anything probative of confusion or no-confusion that doesn't fit the prior 12 factors.
The Two Factors That Almost Always Decide It #
Despite there being 13 factors on paper, the Federal Circuit and TTAB rulings over decades show that two factors dominate the analysis:
- Factor 1 — similarity of the marks themselves
- Factor 2 — similarity of the goods or services
If the marks are very similar AND the goods/services are very similar, you almost always lose. If both differ substantially, you almost always win. The other 11 factors mostly come into play in close cases or when one of the first two is borderline.
Practical implication: when assessing your own brand against an existing mark, those two questions are where 80% of your analytical time should go.
Why USPTO Approval Doesn't Mean You're Safe in Court #
This is the most common — and most expensive — misconception in trademark practice.
There are three distinct forums that apply the likelihood of confusion test, and they don't always reach the same conclusion:
- USPTO examining attorney — pre-registration, ex parte (just you and the examiner). Applies all 13 DuPont factors with limited evidence.
- Trademark Trial and Appeal Board (TTAB) — post-publication opposition or cancellation. Both parties brief, evidence is fuller, but no jury and no damages.
- Federal District Court — infringement lawsuits. Damages, injunctions, juries available. Full discovery, full evidence.
A trademark can register at the USPTO (clean Section 2(d) review) and still get successfully sued in federal court. Why?
- The court sees evidence the USPTO didn't — actual consumer confusion, marketing materials, channels of trade in practice.
- Common law trademark rights — established by use, not registration — can defeat a federal registration in court.
- Famous marks get broader protection in court than narrow class-based USPTO reviews capture.
The takeaway: USPTO approval reduces your risk but does not eliminate it. A clean registration is a green light to launch the brand, not a guarantee that no one will sue you.
The "Three Threats" of Confusion (What You Actually Face) #
Likelihood of confusion shows up in your business in three different ways, each with its own cost profile:
Threat 1: Section 2(d) Office Action #
What it is: A USPTO examining attorney refuses your application because they identify a confusingly similar prior mark.
How common: ~1 in 5 USPTO applications gets a Section 2(d) refusal.
Cost: The $350 per class filing fee is non-refundable. Six months waiting to respond. Optional $500-$3,000 attorney fees to argue back. Often the cheapest outcome of the three threats, but still cash you don't get back.
Threat 2: TTAB Opposition or Cancellation #
What it is: Your mark publishes for opposition (or is already registered), and a third party files an opposition or cancellation proceeding at the Trademark Trial and Appeal Board.
How common: Much rarer than Office Actions but devastating when it happens.
Cost: $5,000 - $50,000+ in attorney fees to defend. 2-3 years to resolution. Brand launch effectively on hold the entire time.
Threat 3: Federal Court Infringement Lawsuit #
What it is: An existing trademark owner sues you in federal court for trademark infringement under the Lanham Act.
How common: Rare in absolute terms, but the existential threat. Doesn't require the plaintiff to be registered — common law rights are enough.
Cost: $50,000 - $500,000+ to defend. Potential damages: lost profits, disgorgement of your profits, attorney fees, treble damages for willful infringement, statutory damages up to $2 million for counterfeiting.
This is why the test matters: every business decision you make about your brand — naming it, designing the logo, picking the class to file — is implicitly a bet on which side of the likelihood-of-confusion line you'll fall.
How to Self-Test Your Own Mark #
Before spending money on filing or branding, run this six-question self-assessment against every potentially-conflicting mark you find in your search:
- Sight test: Print both marks side-by-side. Squint. Do they look like they could come from the same source?
- Sound test: Say both marks out loud at normal speaking speed. Could you mishear one for the other in a noisy room?
- Meaning test: What does each mark connote? Do they evoke the same idea or image?
- Goods test: Are your goods/services in the same class? In related classes? Sold through the same channels?
- Fame test: Is the prior mark famous outside its registered class? (If yes, scope of protection widens.)
- Channel test: Where will consumers encounter your mark? Same retailers? Same platforms? Same price tier?
If two or more of these come back as "yes, similar," you have meaningful likelihood-of-confusion risk and should reconsider — change the mark, change the class, or get formal legal review before filing.
Red Flags: When to Stop and Reconsider #
- You found a mark that's phonetically identical to yours in the same or related class
- You found a famous mark that uses similar wording, even in a totally different class
- You found a mark with similar wording sold through the same retailers you plan to sell through
- The prior mark owner has a track record of opposing similar filings (check TTAB records)
- You found "actual confusion" evidence — customers asking whether the two brands are related, or social media confusion between them
- The conflicting mark spans many product categories — wider use = wider protection
Frequently Asked Questions #
What's the difference between likelihood of confusion at the USPTO vs. in court? #
Same legal test, different evidence and stakes. The USPTO applies the 13 DuPont factors with limited evidence (essentially the marks and goods on paper). A federal court applies the same factors with full discovery, actual marketing evidence, and consumer confusion data. A mark can pass USPTO scrutiny and still lose in court.
Is "likelihood" higher or lower than "more likely than not"? #
Lower. The plaintiff doesn't need to prove that confusion is more likely than not — only that confusion is a probable result of consumers encountering both marks. Courts have rejected attempts to set the bar at "preponderance of confusion."
Does it matter if I didn't know about the other mark? #
For likelihood of confusion analysis, no. Your knowledge or intent doesn't factor into whether consumers would be confused. Intent matters for damages and remedies — willful infringers face treble damages — but a good-faith adopter can still infringe.
Can two identical marks coexist if they're in different industries? #
Often yes, sometimes no. The key is whether the goods or services are "related" in a way that would lead consumers to expect a common source. Software and software services are related. Apparel and accessories are related. Famous marks (Tesla, Apple, Disney) are protected across all industries because their fame creates expectation of expansion.
Who decides likelihood of confusion? #
At the USPTO: an examining attorney during application review. At the TTAB: a panel of three administrative trademark judges. In federal district court: a judge (on motions) and ultimately a jury (at trial) applying the test under the relevant circuit's standards, which usually overlap heavily with DuPont.
If I get a cease & desist letter citing likelihood of confusion, am I cooked? #
Not necessarily. A C&D letter is the sender's opinion that confusion exists. Many C&Ds are overreach. The right response depends on the strength of the senior mark, similarity of the marks and goods, and your evidence of independent creation. Don't ignore it, but don't panic-rebrand either. Get a written legal opinion.
How can I reduce my risk of a likelihood-of-confusion problem? #
Pre-launch search, in this order: (1) USPTO Trademark Search for your name and phonetic variants; (2) visual logo search if you have a design element; (3) common law search via Google and social platforms; (4) attorney clearance opinion if you're investing serious money. Most refusals and lawsuits hit founders who skipped step 1.
Related reading #
- How to Check If a Brand Name Is Trademarked — pre-filing wordmark search guide
- How to Check If Your Logo Is Already Trademarked — visual mark search guide
- What Happens When You Get a Trademark Takedown on Amazon or Etsy — platform-side enforcement
Conclusion #
"Likelihood of confusion" sounds vague, but in practice the analysis collapses to two questions 80% of the time: are the marks similar, and are the goods/services similar? Get those two wrong and you're exposed at the USPTO, the TTAB, and in court — three different forums applying the same test with different stakes.
The good news: the test cuts both ways. If the marks differ in sight/sound/meaning AND the goods/services differ, your risk drops dramatically. The pre-launch search and self-test exist precisely to surface conflicts before they become Office Actions, oppositions, or lawsuits.
Run the test on yourself before someone else does.
External resources #
- USPTO Trademark Search — search the federal register
- USPTO: Responding to Office Actions — what to do after a 2(d) refusal
- Trademark Trial and Appeal Board — TTAB opposition and cancellation procedures
- USPTO TMEP §1207 — examining attorney guidance on likelihood of confusion
Disclaimer: This article provides general information about U.S. trademark law and is not legal advice. Likelihood-of-confusion analysis is fact-specific; the outcome of any particular case depends on its facts. For specific legal questions about a mark, a refusal, a cease & desist letter, or a lawsuit, consult a qualified trademark attorney.
