In re E.I. du Pont de Nemours & Co.: Trademarks

In re E.I. du Pont, 476 F.2d 1357 (CCPA 1973), built the thirteen-factor confusion test that still governs every Section 2(d) refusal and TTAB dispute.

In re E.I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973), decided May 3, 1973, is the most consequential opinion in American trademark prosecution. Chief Judge Howard Markey, writing for the Court of Customs and Patent Appeals, took an unremarkable dispute over the word RALLY on car polish and used it to catalogue the thirteen evidentiary factors that bear on likelihood of confusion under Section 2(d) of the Lanham Act, 15 U.S.C. § 1052(d). More than fifty years later, every office action refusing registration over a prior mark, every Trademark Trial and Appeal Board opposition on confusion grounds, and every Federal Circuit appeal from those decisions runs through the DuPont factors. No other authority is cited as often in registration practice.

The opinion endures for a second reason that is easy to miss behind the famous list. DuPont is also the foundational case on consent agreements and market-division arrangements between trademark owners. The court reversed the refusal precisely because the parties had negotiated a detailed agreement carving up the RALLY market, and it instructed the Patent Office to take such agreements seriously rather than substituting its own speculation about confusion for the judgment of the businesses closest to the marketplace.

At a glance

Case: In re E.I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973)
Decided: May 3, 1973, by the U.S. Court of Customs and Patent Appeals, opinion by Chief Judge Markey; the Board’s refusal of registration was reversed
Holding: Likelihood of confusion under Section 2(d) must be assessed on the facts of each case through thirteen enumerated factors considered when of record, and a well-crafted agreement between the parties allocating their markets can tilt the scales decisively against confusion
Significance: The DuPont factors remain the governing test for confusion in USPTO examination, TTAB proceedings, and Federal Circuit review as of July 2026, embedded in TMEP § 1207.01 and reaffirmed in modern decisions such as In re Charger Ventures LLC, 64 F.4th 1375 (Fed. Cir. 2023)

A refusal, an acquisition, and a divided market

The facts were prosaic. DuPont first applied in 1967 to register RALLY for a combination wax and cleaning agent for automobiles. The examiner refused registration under Section 2(d), citing an existing registration of RALLY for an all-purpose detergent owned by Horizon Industries Corporation, and the Board affirmed. On the face of the register, the case looked easy: identical word, overlapping field of cleaning products.

What happened next made the case. While that appeal was pending, DuPont bought Horizon’s RALLY mark for the automobile product, the good will of that business, and a separate application Horizon had itself filed in 1968 for RALLY on a combination polishing, glazing and cleaning agent for automobiles. That assigned application, not DuPont’s original one, is the case the CCPA decided. Because Horizon kept RALLY for its all-purpose detergent, the companies executed an agreement the same day dividing the field: Horizon’s realm was the commercial building and household market, DuPont’s was the automotive aftermarket, products merely incidentally usable in the other’s market could still be sold, but neither party would promote its products as especially suited for the other’s market. In other words, the two entities with the most intimate knowledge of the products, the customers, and the trade channels had structured their affairs so that confusion would not occur, and each had a commercial incentive to police that line.

The examiner, aware of both the assignment and the agreement, refused the assigned application anyway, describing the issue as already “ruled upon” in the earlier decision. The Trademark Trial and Appeal Board affirmed, holding that “despite any agreement between the parties the public interest cannot be ignored.” The CCPA reversed.

The thirteen factors

Chief Judge Markey’s response to the Board was to insist that Section 2(d) analysis cannot be done by maxim. The statutory question is whether the applied-for mark “so resembles” a prior mark “as to be likely, when used on or in connection with the goods of the applicant, to cause confusion, or to cause mistake, or to deceive.” Answering it, the court held, requires consideration of every probative fact in the record. The opinion then enumerated the factors that, “when of record, must be considered”:

The similarity or dissimilarity of the marks in their entireties as to appearance, sound, connotation, and commercial impression.
The similarity or dissimilarity and nature of the goods or services described in the application and registration.
The similarity or dissimilarity of established, likely-to-continue trade channels.
The conditions under which and buyers to whom sales are made, that is, impulse versus careful, sophisticated purchasing.
The fame of the prior mark, measured by sales, advertising, and length of use.
The number and nature of similar marks in use on similar goods.
The nature and extent of any actual confusion.
The length of time during, and conditions under which, there has been concurrent use without evidence of actual confusion.
The variety of goods on which a mark is or is not used.
The market interface between the applicant and the owner of the prior mark, including mere consents, agreements designed to preclude confusion, assignments of the mark and goodwill, and laches or estoppel.
The extent to which the applicant has a right to exclude others from use of its mark on its goods.
The extent of potential confusion, whether de minimis or substantial.
Any other established fact probative of the effect of use.

The thirteenth factor is a deliberate catch-all. It converts the list from a closed code into an open evidentiary framework, and the TTAB has used it over the decades to absorb everything from prior litigation history to family-of-marks arguments.

No litmus rule

Just as important as the list is the methodology the court attached to it. Markey warned that “each case must be decided on its own facts” and that there is no litmus rule providing a ready guide to all cases. Not every factor is relevant in every proceeding; only those on which evidence appears in the record must be weighed. The significance of any single factor varies from case to case, and any one factor may control a particular dispute.

That flexibility is the test’s great strength and its persistent frustration. It prevents the mechanical outcomes the Board had reached in the RALLY case itself, where identical marks in a loosely related field were treated as the end of the analysis. But it also means that likelihood of confusion is notoriously difficult to predict, since two tribunals weighing the same record can emphasize different factors. The Federal Circuit polices the line mostly through explanation: in In re Charger Ventures LLC, 64 F.4th 1375 (Fed. Cir. 2023), the court affirmed a refusal of SPARK LIVING over SPARK for real estate services even though the Board never announced the weight given to each factor, because the Board had given a reasonable explanation for each factor it considered and its path could reasonably be discerned. Factual findings on each factor are reviewed for substantial evidence; the ultimate legal conclusion of likely confusion is reviewed de novo.

The market interface and the weight of agreements

Applying its own framework, the court found the tenth factor dispositive. The assignment of Horizon’s automotive business and the accompanying market-division agreement were not mere paper consents; they were detailed commercial arrangements by parties whose self-interest ran against confusion. The opinion is memorably practical on this point, observing that it is a matter of economic survival for a businessman to obtain and retain customers and to avoid and preclude confusion, and that millions of advertising dollars are spent daily for that precise purpose. When those most familiar with conditions in the marketplace structure their affairs to avoid confusion, the scales of evidence tilt accordingly, and the Patent Office should hesitate before overriding that judgment with its own armchair prediction.

DuPont thus became the charter for consent-agreement practice in trademark prosecution. Examining attorneys still refuse applications despite consents when an agreement is a “naked” one, reciting permission without explaining why confusion is unlikely. But a clothed consent, one that recites the parties’ market separation, trade-channel differences, and undertakings to avoid overlap, carries substantial weight under DuPont and its progeny, and the TMEP instructs examiners accordingly.

Open questions

Fifty years of application have not closed every seam. The Federal Circuit has never fixed how much articulation the Board owes when it weighs the factors, and Charger Ventures shows the court tolerating fairly thin explanations so long as the path is discernible; where the line falls between discernible and arbitrary remains contested. The relationship between the DuPont framework and the regional circuits’ infringement tests is another live issue: after B&B Hardware, Inc. v. Hargis Industries, Inc., 575 U.S. 138 (2015), a TTAB determination made under DuPont can preclude relitigation of confusion in an infringement suit when the Board’s usages match the marketplace usages, which raises the stakes of registration proceedings in ways the 1973 court never contemplated. And the weight of consent agreements continues to generate friction, since the USPTO periodically signals skepticism of consents in crowded fields while applicants read DuPont as a command of deference.

Implications for brands and businesses

Build the record around the factors that favor you. Only factors “of record” must be considered, so an applicant facing a Section 2(d) refusal should submit evidence, not just argument: third-party registrations to show a crowded field under factor six, declarations on trade channels and purchaser sophistication under factors three and four, and years of concurrent use without incident under factor eight.
A well-drafted consent agreement is the most powerful single document in confusion practice. DuPont itself was won on the market interface. A consent that recites concrete confusion-avoidance measures, market divisions, and the parties’ business judgment stands a real chance of overcoming a refusal; a naked consent usually does not.
Do not assume identical marks mean an automatic loss. The whole point of DuPont is that confusion is a fact question spanning thirteen dimensions. Identical words can coexist where goods, channels, and buyers genuinely diverge, and conversely, dissimilar goods can still conflict where a prior mark is famous.
Treat TTAB proceedings as potentially outcome-determinative. Because DuPont findings can have preclusive effect in later litigation under B&B Hardware, a business should litigate an opposition or cancellation with the same rigor it would bring to district court, or consider whether registration-level compromise is the wiser course.

Frequently asked questions

Do all thirteen DuPont factors have to be analyzed in every case? No. The court said the factors must be considered “when of record,” meaning only the factors on which the parties actually submitted evidence come into play. In most ex parte examinations only a handful are litigated, usually similarity of the marks and relatedness of the goods or services, and the CCPA was explicit that any single factor may control a given case.

Does DuPont apply in infringement litigation too? Not directly. DuPont governs registration decisions at the USPTO, the TTAB, and the Federal Circuit. The regional circuits apply their own multifactor confusion tests in infringement suits, such as the Second Circuit’s Polaroid factors and the Ninth Circuit’s Sleekcraft factors. The frameworks overlap heavily, and after B&B Hardware, Inc. v. Hargis Industries, Inc., 575 U.S. 138 (2015), a TTAB confusion ruling can even preclude relitigation in court when the usages at issue are materially the same.

Why did DuPont win when both marks were the identical word RALLY? Because of the tenth factor, the market interface. DuPont had purchased Horizon’s RALLY mark for the automobile product, the application at issue, and the goodwill of that business, and the parties signed an agreement dividing the markets: Horizon kept commercial building and household cleaning, DuPont took the automotive aftermarket. The court held that a detailed agreement between sophisticated businesses, each motivated to avoid confusion, was weighty evidence that confusion was unlikely.