Trademark likelihood of confusion is the legal test that asks whether an ordinary consumer, shopping with normal care, would probably believe two brands come from the same source or are somehow connected. It is the central question in almost every trademark dispute, whether the U.S. Patent and Trademark Office is deciding to register a new mark or a federal court is deciding an infringement lawsuit. The answer turns on a flexible list of factors, and no single factor decides the case on its own.
What the Test Actually Asks
The standard is measured through the eyes of an ordinary, reasonably careful consumer, not an expert with both marks laid side by side.1Ninth Circuit District & Bankruptcy Courts. 15.18 Infringement – Likelihood of Confusion – Factors – Sleekcraft Test The question is whether someone encountering the marks in the real world, with imperfect memory and divided attention, would think the goods share a common source, sponsorship, or affiliation. A careful comparison that reveals subtle differences does not settle it. Real shoppers do not do careful comparisons.
Three provisions of the Lanham Act put this test to work. The USPTO can refuse registration when a new mark too closely resembles an existing one.2Office of the Law Revision Counsel. 15 USC 1052 – Trademarks Registrable on Principal Register Owners of registered marks can sue anyone whose use in commerce is likely to confuse consumers.3Office of the Law Revision Counsel. 15 USC 1114 – Remedies; Infringement; Innocent Infringement by Printers and Publishers Even unregistered marks get protection against confusingly similar uses.4Office of the Law Revision Counsel. 15 USC 1125 – False Designations of Origin, False Descriptions, and Dilution Forbidden The likelihood-of-confusion analysis is the same in each setting.
The Factors Courts Weigh
The framework traces to the 1973 decision in In re E.I. du Pont de Nemours & Co., which set out thirteen factors the USPTO and courts consider when the evidence is available.5TransJurLex. In re E.I. duPont de Nemours and Co., 476 F.2d 1357 (CCPA 1973) Federal circuits use their own variations. The Ninth Circuit calls its version the Sleekcraft test; the Third Circuit uses the Lapp test. The underlying inquiry is the same everywhere: compare the marks, compare the goods, and assess how real consumers would react.
No single factor controls. Identical goods can still coexist under marks that look and sound entirely different. Marks that share only a vague resemblance can still infringe when used on identical products sold to the same audience. In most cases, the similarity of the marks and the relatedness of the goods carry the most weight, but everything else is on the table.
Similarity of the Marks
Examiners and judges compare marks along four dimensions: appearance, sound, meaning, and overall commercial impression.6United States Patent and Trademark Office. Likelihood of Confusion Marks do not need to be identical for confusion to be likely. Two names that share a similar look, a similar pronunciation, or a similar underlying concept can create enough overlap.
Visual similarity covers letterforms, color schemes, stylization, and logo design. Sound similarity matters because consumers often hear a brand name in conversation or in an audio ad without ever seeing the spelling. Two names that look different on paper but sound almost the same when spoken can trip up a buyer. Meaning similarity catches marks that use different words to convey the same idea. Under the doctrine of foreign equivalents, a foreign-language word familiar to an appreciable segment of American consumers can be treated as equivalent to its English translation, so “Sol” and “Sun” for competing products may end up in the same territory.
The analysis looks at the marks as a whole, not dissected into parts. A shared prefix, a common suffix, or a similar rhythm can drive the overall impression even when other elements differ. And because consumers rarely see both marks together, the comparison happens against a fuzzy mental image. Modest similarities are often enough.
Relatedness of the Goods or Services
Two marks can coexist peacefully when one sells jet engines and the other sells cupcakes. The risk rises when the products are commercially connected enough that a buyer encountering them under similar marks would assume a single source is behind both.7United States Patent and Trademark Office. Possible Grounds for Refusal of a Mark
Complementary goods are a frequent flashpoint. Computer hardware and software, wine and glassware, athletic shoes and sports apparel are all pairs consumers naturally associate. Trade channels matter too. If both products sit on the same shelf, appear in the same online marketplace, or target the same demographic, the overlap strengthens the confusion argument. A premium skincare line sold only through dermatologists faces a different analysis than one sold at a mass-market drugstore, even under somewhat similar marks.
Strength of the Senior Mark
The stronger and more distinctive the original mark, the wider the zone of protection it commands. Trademark law places marks on a spectrum, and where a mark falls on that spectrum directly affects how easily the owner can prove confusion.
- Fanciful marks are invented words with no meaning outside the brand, like XEROX or KODAK. They receive the broadest protection.
- Arbitrary marks use real words in unrelated contexts, like APPLE for computers. They get the same high level of protection.
- Suggestive marks hint at a quality of the product but require the consumer to make a mental leap, such as GREYHOUND for bus service. They are inherently distinctive and protectable without extra proof.8Legal Information Institute. Distinctive Trademark
- Descriptive marks directly describe a feature or quality of the product, like “Cold and Creamy” for ice cream. They earn protection only when the owner can show the public has come to associate the term with a single source, a concept called secondary meaning.9Legal Information Institute. Secondary Meaning
- Generic terms, like “computer” for computers, can never function as trademarks no matter how much advertising money goes behind them.
A fanciful or arbitrary mark with high consumer recognition creates a formidable barrier for anyone filing a remotely similar mark, even for loosely related goods. A descriptive mark with modest secondary meaning has a much narrower scope and may only block near-identical marks on highly overlapping goods.
Buyer Sophistication and the Junior User’s Intent
The care a typical buyer exercises when purchasing the goods matters. Expensive, specialized products attract more scrutiny. A hospital purchasing committee evaluating a $200,000 piece of medical equipment will compare brands far more carefully than a grocery shopper grabbing a $3 bag of chips. The more sophisticated and deliberate the buying process, the less likely similar marks will actually confuse anyone.1Ninth Circuit District & Bankruptcy Courts. 15.18 Infringement – Likelihood of Confusion – Factors – Sleekcraft Test Impulse purchases at low price points push the analysis the other way, because consumers in that setting simply are not paying close attention.
The junior user’s intent also plays a role. When evidence shows a newcomer deliberately chose a mark to trade on the goodwill of an established brand, courts treat that as strong circumstantial proof that confusion is likely. The reasoning is practical: if the infringer believed the similarity would attract the senior user’s customers, that belief is probably correct. Intent alone does not decide the case, but it can tip the balance when other factors are close.
Evidence of Actual Confusion
Nothing strengthens a confusion claim like proof that real consumers have already been misled. Misdirected emails, phone calls meant for the other company, and social media posts tagging the wrong brand all qualify. This kind of evidence is not required to win, but when it exists it is enormously persuasive. It turns the analysis from a prediction about what might happen into a record of what already has.
Attorneys sometimes commission formal consumer surveys to generate statistical evidence. A well-designed survey following accepted methodology, such as the Eveready or Squirt formats, can quantify the percentage of consumers who associate one mark with the other. These surveys are expensive and the results are frequently contested, but courts give weight to properly conducted studies. The reverse is also true. A long period of coexistence in the market without any evidence of actual confusion can undercut a claim, particularly when both marks have been in heavy use.
Two Variations Worth Knowing
Reverse Confusion
Most infringement cases involve a smaller newcomer copying a bigger, established brand. Reverse confusion flips that dynamic. It happens when a larger company adopts a mark that a smaller, less well-known business already owns, and then pours so much marketing muscle behind it that consumers start associating the mark with the larger company. The smaller senior user ends up looking like the knockoff of its own brand. Courts recognize this as infringement because it effectively strips the smaller company of its trademark identity, even when nobody set out to copy anyone.
Initial Interest Confusion
Confusion does not have to result in a completed sale to count. The initial interest confusion doctrine recognizes infringement when a consumer is momentarily drawn to a competitor’s product because of a similar mark, even if the consumer figures out the mistake before buying anything. This comes up frequently in digital contexts, such as a company bidding on a competitor’s trademark as a search engine keyword. The diversion of attention itself is treated as a misappropriation of goodwill, whether or not money changes hands.
If the USPTO Cites Confusion Against Your Application
A likelihood-of-confusion refusal is one of the most common grounds the USPTO gives for rejecting a trademark application, and it does not mean your application is dead.7United States Patent and Trademark Office. Possible Grounds for Refusal of a Mark You have several options for responding.
The most direct is to argue the differences. Make the case that the marks create distinct commercial impressions, that the goods target different markets, or that the trade channels do not overlap. Supporting evidence might include specimens showing how the marks actually appear in the marketplace, evidence of different consumer demographics, or proof that the cited mark is commercially weak because many similar marks already coexist on the register for related goods.
Another option is narrowing your application. If the overlap exists only because your goods description is broadly worded, amending it to exclude the conflicting category can sometimes resolve the refusal without giving up anything you actually sell. Finally, if you can reach the owner of the cited mark, a consent agreement in which both parties acknowledge the coexistence and agree to measures preventing confusion can persuade the USPTO to withdraw the refusal. A trademark attorney familiar with office action responses is worth the investment, because the arguments that work are highly fact-specific and the wrong strategy can create problems that follow the mark for years.