We parsed 174,986 real USPTO Office Action responses to see which TTAB and Federal Circuit cases practitioners reach for most — and how often the applications that cited them actually registered. The answer is not what the citation counts suggest.
A reference library of all 15 landmark cases — what each held, the verbatim holding, the registration rate of the applications that cited it, and how practitioners cite it.
Editor’s note — How to read these numbers
These rates describe what happened to applications that cited each case — not what the case did for them. Citing a precedent does not win a response, and a low rate does not make a case weak. Often the opposite: the hardest refusals attract the most-cited authorities, which drags their measured rate down.
Read each number against its refusal-ground baseline — §2(d) likelihood-of-confusion cases clear at a very different rate than §2(e)(1) descriptiveness cases. The right comparison is a case versus its own ground, never §2(d) against §2(e)(1).
The 15 cases practitioners cite most against §2(d) and §2(e) refusals — grouped by ground and sortable by registration rate. Each links to a full breakdown with the verbatim holding and a model citation paragraph.
Least-sophisticated purchaser
42 of 164 citing apps terminal
Run a §2(d) clearance analysisForeign equivalents & fame
139 of 518 citing apps terminal
Run a §2(d) clearance analysisDominant feature; disclaimed matter
56 of 194 citing apps terminal
Run a §2(d) clearance analysisFirst-word dominance; similarity
32 of 172 citing apps terminal
Run a §2(d) clearance analysisSimilarity; standard-character display
38 of 211 citing apps terminal
Run a §2(d) clearance analysisFame as a §2(d) factor
91 of 350 citing apps terminal
Run a §2(d) clearance analysisMarks compared in their entireties
190 of 603 citing apps terminal
Run a §2(d) clearance analysisCrowded field; weak design elements
105 of 432 citing apps terminal
Surface crowded-field evidenceThe §2(d) factors
455 of 1,740 citing apps terminal
Run a §2(d) clearance analysisRelatedness of goods and services
52 of 195 citing apps terminal
See coexistence evidence on the registerDescriptiveness standard
47 of 215 citing apps terminal
Draft a descriptiveness responseCorpus. 174,986 distinct parsed USPTO Office Action responses (snapshot: June 2026). For each precedent we phrase-matched the response text, then scoped to the responses whose refusal ground matches the case's doctrine, counted at the application (distinct serial) level.
Rates. "Registered" is the share of terminal applications — those that have reached a final registered or abandoned outcome — that registered. Pending applications are excluded, so recent, heavily-cited cases rest on a smaller terminal base.
Correlation, not causation. A response can cite a precedent and also face an unrelated refusal; a low rate reflects that hard refusals attract case-law briefing, not that a citation hurts. Nothing here should be read as "cite this case to win."
Holdings. Each summary is drawn from the cited primary opinion — the TTAB e-FOIA reading room for TTAB decisions, and the Federal Circuit / CCPA opinions for the appellate cases. All are public-domain government works.
Precision. A few case names still rely on loose phrase-matching rather than a curated set of reporter citations, which can modestly affect the smallest counts.
From precedent to filing