Treat legacy Paper D as a regime-switching exam: for every task, first decide whether the EPC, national enforcement law, or the PCT governs the issue, then answer with a stated legal requirement applied to the stated facts. The sections below build that habit with two worked scenarios, a legal-basis table, and a self-check rubric you can run against every practice answer.
Separating validity analysis from infringement analysis in the legal opinion
A Paper D opinion asks two questions about one patent: is it valid under the EPC, and do the described acts infringe? These are governed by different frameworks, so keep them in separate sections with separate legal sources.
The EPC is a grant treaty: it sets uniform rules for obtaining and maintaining European patents through the EPO. Article 64 EPC describes the rights a granted patent confers, but enforcement of those rights rests with the courts of the contracting states and, where opted into, the Unified Patent Court. An opinion that tests infringement by citing EPC patentability articles has mixed up the frameworks; novelty and inventive step belong in the validity section, not the infringement section.
A workable structure gives each client question its own heading: claim construction, infringement per accused act, validity per ground, and a concluding recommendation. Under each heading, state the legal requirement in one or two sentences, apply the facts, and conclude with a confidence level, flagging factual uncertainties such as whether the accused product truly carries every claimed feature. Conclusions without a stated requirement are hard to follow and easy to contradict.
Scenario 1: the infringement opinion aimed at the wrong forum
In this scenario a candidate advises suing 'at the EPO' for infringement. The better decision identifies which national or regional court can hear the case, because the EPO grants patents but does not adjudicate infringement.
Facts: a client holds a European patent validated in several states and learns a competitor is selling an identical product across those markets. A plausible mistake is to write the whole opinion as if a single European infringement court existed by default, quoting EPC articles on patentability to support an infringement claim. That answer loses the thread: it identifies no actionable forum and uses the EPC for something it does not decide.
The better opinion first checks whether each accused act falls within the claims, then explains that infringement proceedings run before the national courts of the states concerned, or the Unified Patent Court where the patent falls under it, each applying its own interpretation emphasis. It then weighs practical factors: where the commercial harm concentrates, whether one forum can cover several states, and whether an EPO opposition against a competitor's patent is the sensible parallel move. Forum choice changes everything downstream.
Reading claims at the EPO versus reading claims in court
Claim interpretation differs by context. During EPO proceedings claims are read for clarity and patentability; in enforcement, Article 69 EPC and its Protocol set the framework, with the description interpreting terms rather than erasing them.
In examination and opposition, the operative question is what a skilled reader understands the claim to cover, and whether that coverage is clear, supported and allowable. The description supplies the meaning of technical terms, but features absent from a claim cannot be imported into it, and examples do not narrow a claim that reads beyond them. Practise writing a one-sentence construction for each disputed claim term before arguing novelty or inventive step, because both analyses inherit that construction.
In infringement settings, the Protocol on Article 69 guards against two opposite errors: reading claims as literal recipes stripped of context, and treating them as mere guidelines. Because national courts differ in emphasis, an opinion should state its construction, note that conclusions are framed under that construction, and avoid asserting a uniform European infringement outcome. The same patent can support different infringement conclusions in different states; the opinion should say so rather than paper over the difference.
Scenario 2: amending claims without falling into added-matter traps
A candidate answers an inventive-step objection by importing features drawn from two different embodiments of the description. The better amendment uses a single supported embodiment and is checked against both Article 123(2) and Article 123(3).
Facts: an examiner objects that the independent claim lacks inventive step; the description discloses embodiment A with a mechanical feature and embodiment B with a thermal feature, presented as alternatives. A tempting amendment combines both features 'to be safe'. That combination is arguably not directly and unambiguously derivable from the application as filed, since the filing never taught the combination, so it invites an added-matter objection under Article 123(2) and a fresh clarity fight.
The better decision selects one embodiment, amends with its feature only, and verifies support sentence by sentence against the filing, noting the passage relied on. In opposition, the same amendment must also pass Article 123(3): the amended claim may not cover subject matter the granted claim did not, so deleting features to widen scope is checked, never assumed. Keeping the cited support passages in the response makes the amendment auditable for the reader.
Choosing the right legal basis: a decision table for attack and amendment
Every argument in Paper D sits on a named provision. Match the situation to the provision before writing: novelty, inventive step, priority, added matter, scope extension and the PCT's status each trigger different checks.
Distinguish the two validity attacks structurally. Novelty under Article 54 EPC compares the claim against one single prior disclosure and asks whether every feature is found there; a near miss is fatal to the novelty attack but irrelevant to inventive step. Inventive step under Article 56 asks whether the skilled person, starting from one document and faced with the objective technical problem, would have arrived at the claim. Label which document starts the analysis and where each feature comes from.
The same discipline applies when amending and when a PCT question appears. Added matter and scope extension are tests about the application's own text and the granted text, not about prior art. A PCT application is an international filing procedure; it never matures into an enforceable right by itself, so advice must identify national or regional phase entry as the step that produces rights. Mixing these categories produces confident-sounding but misdirected answers.
| Situation | Governing provision | Core question | Verification check |
|---|---|---|---|
| Attacking a claim against one document | Art. 54 EPC (novelty) | Does one prior disclosure show every claimed feature? | Feature-by-feature chart; one missing feature ends the attack |
| Arguing a claim is obvious | Art. 56 EPC (inventive step) | Would the skilled person arrive at the claim from the closest prior art and the objective problem? | Name closest prior art, difference, technical effect, problem |
| Amending during examination or opposition | Art. 123(2) EPC (added matter) | Is the amendment directly and unambiguously derivable from the application as filed? | Cite the supporting passage for each added feature |
| Changing a granted claim in opposition | Art. 123(3) EPC (scope) | Does the amended claim extend beyond the granted protection? | Compare old and new claim scope expressly |
| Advising on a PCT application | PCT filing procedure | Where and when must national or regional phase steps be taken to obtain rights? | Identify the pending phase; do not treat the PCT application as a granted right |
Procedure, admissibility and representation: checks that decide whether merits matter
A merits argument cannot rescue an inadmissible request. Paper D-style tasks expect you to check representation, match the client's goal to the correct procedure, and confirm formal requirements before substantively arguing anything.
Before the EPO, a party without domicile or seat in a contracting state generally needs a professional representative entered on the list maintained under Article 134 EPC, so a client letter should flag representation needs rather than assume them. For each client goal, match the procedure: a granted patent can be opposed within the opposition period, attacked later through national revocation actions, or narrowed through the EPO's limitation procedure; appeals lie against certain first-instance decisions, not against everything.
Run admissibility checks before merits in your answer and announce them as such: is the request filed by the right party, in time, in the correct form, and does the decision at hand support an appeal at all? Time limits and remedy rules are exactly where administrative detail changes over time, so for the current session verify periods and formalities on the EPO's own pages rather than trusting any summary, including this one.
A preparation sequence and self-check rubric for the final session
Prepare by regime, not by topic list: build the map, drill two full scenario opinions, drill amendments with support checks, then rehearse client letters. Close with a rubric so every practice answer is scored on observable criteria.
A realistic sequence: first, map the frameworks on one page — EPC grant articles, the PCT's procedural role, and where enforcement lives. Second, write two complete legal opinions from published materials or self-built fact patterns, using the structure from Section 1. Third, take three claim sets and practise amendments with cited support passages. Fourth, convert one opinion into a client letter, cutting legal citations down to what a client needs in order to decide.
Score each practice piece against a rubric: (1) correct regime identified for every issue; (2) legal requirement stated before application; (3) conclusion follows from the stated facts, with uncertainty flagged; (4) every amendment carries a cited support passage and an explicit scope comparison; (5) client-facing text recommends an action, not a treatise. If a piece misses two or more criteria, redo it before reading new material. A rubric score is a learning milestone, not a prediction of any exam outcome.
- Readiness check 1: given any issue in a fact pattern, you can state within seconds whether the EPC, national enforcement law, or the PCT governs it.
- Readiness check 2: you can write a two-line legal requirement plus application for novelty and inventive step without re-reading a reference.
- Readiness check 3: every amendment you draft carries a cited support passage and an explicit Article 123(3) scope comparison.
- Readiness check 4: your client letters recommend one action with one alternative and name the key risk in plain sentences.
- Readiness check 5: rubric score of at least four of five on two consecutive practice opinions.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.