EPO vs USPTO Patent Practice: A 2026 Guide

US patent attorneys and in-house counsel who extend a US application into Europe quickly discover that the European Patent Office (EPO) and the United States Patent and Trademark Office (USPTO) operate on different assumptions. The statutes look similar — novelty, inventive step (non-obviousness), enablement, eligible subject matter — but the day-to-day examination practice diverges in ways that decide whether claims survive, and in what scope. This guide focuses on the three areas where the gap hurts US filers the most: added matter, the problem-solution approach to inventive step, and computer-implemented invention (CII) eligibility.

Added matter is the single biggest reason a European patent looks narrower than its US counterpart. Article 123(2) EPC prohibits any amendment that adds subject matter beyond the content of the application as filed. The EPO applies a strict 'gold standard' — an amendment is only allowed if a skilled person can derive it directly and unambiguously from the application as filed, using common general knowledge. Intermediate generalisations, picking one feature out of a combination disclosed only together, or rewriting a claim by combining separately disclosed embodiments will typically fail. Once granted, an added-matter defect is fatal: Article 123(3) EPC also forbids broadening after grant, creating the well-known 'inescapable trap' where a claim cannot be repaired without either adding matter or broadening scope.

USPTO practice on written description under 35 U.S.C. § 112 is more forgiving. US examiners routinely accept amendments that would be refused at the EPO — combining features from different embodiments, generalising a specific example, or importing a numerical range from a table into the claims. A US application drafted with EPO practice in mind looks different: fallback positions are written out explicitly, ranges are disclosed at multiple levels of generality, and each optional feature is described both in combination and in isolation. Filing the same specification at the EPO that comfortably supports broad claim amendments at the USPTO is a common and expensive mistake.

Inventive step at the EPO is decided under the problem-solution approach, a structured three-step test. The examiner identifies the closest prior art, formulates the objective technical problem solved over that prior art, and asks whether the claimed solution would have been obvious to a skilled person starting from the closest prior art and facing that problem. The test rewards structured argument: any effect that supports inventive step must be technical, derivable from the application as filed, and causally linked to the distinguishing features. Commercial success, long-felt need or copying — familiar secondary considerations at the USPTO — carry very little weight at the EPO.

USPTO non-obviousness under 35 U.S.C. § 103 uses the Graham factors and, since KSR, a more flexible 'obvious to try' and common-sense reasoning framework. US arguments often lean on unexpected results affidavits, secondary considerations and expert declarations. Translated directly into an EPO reply, those arguments frequently miss the mark: if the alleged effect is not disclosed in the application as filed and cannot be plausibly derived from it, the EPO will not use it. US applicants should draft the specification with the problem-solution approach in mind — explicitly identifying the technical problem, the technical effect of each distinguishing feature, and comparative data where possible — rather than relying on post-filing declarations to rescue the case.

Computer-implemented inventions expose the deepest doctrinal gap. Under Article 52(2) EPC, programs for computers 'as such' are excluded from patentability, but the EPO grants CII claims routinely when the invention has a technical character. The EPO's approach — the COMVIK line of case law — first checks whether the claim as a whole has technical character (a mixed claim reciting a computer usually clears this hurdle), then assesses inventive step considering only those features that contribute to a technical effect. Non-technical features (business methods, presentation of information, pure mathematics) are folded into the problem as constraints given to the skilled person, not as contributions to inventive step. The winning strategy at the EPO is to identify and claim the technical effect: reduced memory use, lower latency, improved signal processing, control of a physical process.

USPTO eligibility under 35 U.S.C. § 101 uses the Alice/Mayo two-step framework: is the claim directed to a judicial exception (abstract idea, law of nature, natural phenomenon), and if so, does it recite significantly more? US claims that survive Alice by reciting an improvement to computer functionality often map well onto the EPO's technical effect requirement. But claims drafted around business-method or user-interface arguments that work at the USPTO can fail at the EPO for lack of technical character or lack of inventive step. Conversely, claims that pass the EPO's technical test can still fall to Alice at the USPTO if the technical improvement is not clearly articulated in the specification. A single specification supporting both jurisdictions needs to describe the technical problem, the technical solution and the technical effect in engineering terms, not business terms.

Beyond these three areas, several practice differences compound the divergence. The EPO uses a 'whole contents' approach to novelty (Article 54(3) EPC), meaning earlier-filed but later-published European applications count for novelty but not inventive step, and there is no US-style grace period for the applicant's own disclosures beyond a very narrow six-month window under Article 55 EPC. Unity of invention at the EPO is stricter and more expensive to work around than USPTO restriction practice — additional search fees are payable for each further invention searched, and divisional filings must be made while the parent is still pending. Claim fees kick in from the 16th claim and rise sharply from the 51st claim, so US-style claim sets of 40+ claims are rarely cost-effective in Europe without pruning.

The practical drafting playbook for US filers is straightforward. Write the specification with EPO practice in mind from day one, even if the first filing is at the USPTO. Provide explicit fallback positions and intermediate generalisations so that later amendments can survive Article 123(2). Frame the invention around a technical problem and a technical effect, with data where available. For CII cases, describe the technical implementation and the technical effect in engineering terms — do not rely on the business advantage. Keep the independent claims tight, provide layered dependent claims, and plan for a leaner claim set on European regional-phase entry. See the [European Patents service page](/european-patents) for how this maps to a full European filing strategy, and the [how to file a European patent from the US guide](/blog/file-european-patent-from-us) for the regional-phase mechanics.

For US patent counsel supporting European filings, the right European Patent Attorney is the person who can bridge these differences early — reviewing the US draft before filing, flagging added-matter risk, mapping the invention onto the problem-solution framework and, in CII cases, articulating the technical effect in EPO-acceptable terms. The [how to choose a European Patent Attorney guide](/blog/how-to-choose-a-european-patent-attorney) covers the questions to ask. If you would like a pre-filing review of a US application intended for European regional-phase entry, contact Bauer IP through the [European Patent Attorney service page](/services/european-patent-attorney) or at bauerip.eu.