EU Design Invalidity Success Rates in 2026: What 4,600 Decisions Show
The only invalidation rate in public circulation is six years old. Here is the current figure from the full EUIPO corpus — 69.3% on the merits — as the design reform's new invalidity regime arrives.
Ask how often an EUIPO Registered Community Design invalidity action actually succeeds and you will find one number in circulation — and it is six years old. Reddie & Grose's widely-cited quarterly reviews put the invalidation rate at roughly 63% (late 2019) and 74% (mid-2020), then the series stopped. Since then, nobody has published a current figure. With the EU Design Reform's front-loaded, fast-track invalidity regime arriving on 1 July 2026, the base rate it operates on is worth knowing. So we computed it across the full corpus: 4,665 published EUIPO design invalidity and appeal decisions, 2018 to May 2026.
The current base rate: 69%, not 63%
On a merits-only denominator — excluding inadmissible, withdrawn and discontinued matters — 69.3% of first-instance invalidity decisions invalidate the contested design(2,289 invalidations out of 3,305 decisions on the merits). That is squarely between Reddie & Grose's two old data points, which is reassuring: their methodology and ours, six years apart, describe the same broadly applicant-favourable picture. Year to year the rate sits in a 63%–78% band, with a step up around 2021 that has held since.
The presumption of validity is rebuttable — and on the merits, most of the time, it is rebutted.
The reason is structural. EUIPO designs are registered without substantive examination, so the register carries unexamined rights that an invalidity applicant can attack on novelty and individual character. The high success rate also reflects heavy case selection: only a few hundred actions are filed each year against a backdrop of well over 100,000 design applications, so the cases that reach a decision are already strong. Recent General Court rulings reinforce how exacting the standard is — in Crocs (T-228/25, May 2026) a heel strap was too minor a variation to give the clog individual character, and in Deity Shoes (C-323/24, December 2025) the CJEU confirmed there is no originality threshold to clear at all.
First instance convicts; the Board of Appeal is more cautious
High first-instance conviction does not carry through to appeal. Across substantive Board of Appeal decisions, the panel reverses the first instance in whole or part about 24% of the time — and that average hides a striking spread by rapporteur, which we examine separately. As the reform front-loads evidence at first instance, where a case is won or lost shifts earlier — making both the base rate and the appellate odds more decision-relevant, not less.
What moves the rate
The headline average is the least interesting number in the dataset. The rate moves sharply with the type of prior art cited, with the representative instructed (win rates among the most active applicant firms vary widely), and by Locarno sector. China is the single largest source of first-instance filings, ahead of Germany — a consequence of the requirement that non-EEA parties appoint an EEA representative before the EUIPO.
Method
Figures cover every published EUIPO RCD invalidity and design Board of Appeal decision, 2018 to May 2026, with each decision's text read into a structured record and validated by sampling (98.8% coverage). The merits-only denominator excludes non-merits dispositions; full methodology and cross-checks against EUIPO's own published appeal statistics are in the free annual report, or explore the live figures in the decision analytics.