Share this article:

Unity and The EPO – A Clarified Approach?


In March 2021 the EPO published its revised Guidelines for Examination (hereafter, the “Guidelines”). One area of note that received extensive amendments was Part F Chapter V – Unity. In this article we take some time to examine what has changed, and how this might affect applicants.

A Brief Review

According to the EPC, a European patent application shall relate to one invention only or to a group of inventions so linked as to form a single general inventive concept (Article 82 EPC). Thus, our unity article introduces the notion of a “single general inventive concept”. The rules elaborate: where a group of inventions is claimed in a European patent application, the requirement of unity of invention under Article 82 shall be fulfilled only when there is a technical relationship among those inventions involving one or more of the same or corresponding special technical features. The expression “special technical features” shall mean those features which define a contribution which each of the claimed inventions considered as a whole makes over the prior art (Rule 44(1) EPC).

Thus, while the related Rule 44 does not provide much clarity on what a “single general inventive concept” may mean, it introduces the concept of a “special technical feature.” Nevertheless, it is implied by the Rule that claims having the “same or corresponding” special technical feature(s) meet the requirement of unity. A “special technical feature” must define a contribution over the prior art and, therefore, in order to be considered “special and technical,” a given feature must be novel and inventive over the prior art.

If this needed affirming (the word “inventive” is used in the text of Article 82 EPC), the Guidelines do so. From the previous version of the Guidelines: The term “special” means that the features in question define the contribution that the invention considered as a whole makes over the prior art in terms of novelty and inventive step (F-V, 2, “Requirement of unity of invention”).

Was There a Need for a Change in the Guidelines?

In this author’s opinion, yes. The previous version of the Guidelines could be taken to imply that the presence of a special technical feature was necessary to establish a lack of unity, just as it was necessary to establish unity. In other words – where two claims both had special technical features but these were different, then a lack of unity could be alleged.

The problem this could cause (and, in practice, has) is best illustrated by way of an example, which is based on a real case:

Independent claim 1 has dependent claims 2-15. There are no other dependent claims. In a partial search opinion, a lack of unity was alleged. The following inventions were recognised:

Invention I: claims 2, 10

Invention II: claim 3

Invention III: claims 4, 8, 9

Invention IV: claims 5-7, and 11-15

The reason for the dividing-up and grouping of the claims was as follows. Claim 1 was deemed to lack novelty. The common matter amongst all of the dependent claims – the subject matter of claim 1 – was therefore not considered novel. A lack of unity was therefore deemed present among the dependent claims because they did not contain common, novel and inventive, subject matter (e.g. the dependent claims did not contain a common special technical feature).

The reasons for the particular (and, at first glance, rather odd) grouping above that was adopted by the Examiner was that different “special technical features” were recognised in, and were common to, each dependent claim listed in the above respective groups.

So, where is the problem?

In the Search Opinion, when recognising each of the above four groups of inventions, the Examiner stated that the claims in each group comprised a (different) special technical feature. That they were different was used to form the grouping. “Inventions” I and III (claims 2, 4, 8, 9, and 10) were all the subject of a formal search[1] and, in the Search Opinion, each of the searched claims was the subject of a novelty or inventive step objection.

There was, therefore, a somewhat paradoxical statement by the Examiner: how can two claims contain different special technical features and yet face novelty/inventive step objections? Put another way, EPO jurisprudence, as clarified by the Guidelines at least, is clear in that a “special technical feature” must be novel and inventive (otherwise how could it “make a contribution over the prior art” in the sense of Rule 44 EPC). So, then, how can a claim that admittedly comprises a special technical feature be deemed to lack novelty?

This is not the only case encountered by the author where such a seemingly paradoxical argument has been made by an Examiner, and other attorneys at this firm have reported similar experiences.

As a side note, once groups of dependent claims have been the subject of a unity objection, as in the above example, it can be difficult if not impossible to persuade the Examiner that they should remain within the claim set. For example, even following an amendment introducing novel and inventive subject matter in claim 1 (as an attempt to unify the dependent claims – they would then have a novel and inventive, and therefore a “special technical”, feature in common), the unity objection can linger over the dependent claims. This can manifest itself in the form of an objection to the dependent claims relating to unsearched subject matter (if only a subset of the dependent claims have been deemed searched) and/or a request to cancel dependent claims relating to all but one invention (if the Examiner maintains that there is non-unity among the dependent claims), essentially requiring cancellation of some claims in either instance.

Therefore, following such a unity objection that divides-up dependent claims, even where claim 1 has been amended to recite a special technical feature, it can difficult to successfully argue that some dependent claims should not have to be cancelled.

What are the Changes to the Guidelines?

First of all, F, V-2 (“Requirement of unity of invention”) has been amended to emphasise that a group of inventions may be formed by a plurality of dependent claims where the independent claim is either not novel or not inventive. This is certainly a clarification of the kind of argument that can be raised by the Examiner as illustrated above.

The Guidelines further clarify the meaning of the term “special.” The same passage cited in the first section of this article has been amended, and now reads: The term “special” means that the features in question define the contribution that the invention considered as a whole makes over the “prior art at hand” in terms of novelty and inventive step. The notable change to this passage is that, instead of making a contribution over the prior art, the term “special” refers to making a contribution over the “prior art at hand,” which the new Guidelines define as: “i.e. the prior art relied upon in the non-unity assessment.” Since the prior art at hand may change during the course of proceedings, an assessment of unity is therefore iterative (see also F-V, 3).

In other words, the new Guidelines emphasise that the prior art at hand, on which the assessment of unity is based, is the prior art that is available “at the time” and which may change during the prosecution. We consider this to be a welcome clarification since, in a broad sense, it emphasises that an Examiners original (e.g. at the time of Search) unity assessment should be subject to change, possibly in view of amendments made to “unify” the claims.

Extensive amendments have been made to one section of the Guidelines (F-V, 3, “Assessment of unity”) that, it appears, are intended to clarify the procedure to which Examiners should be adhering when raising a unity objection.

This section now recites that a unity assessment is supposed to determine if the subject-matter of the claims have anything in common that represents a single general inventive concept.

To achieve this aim, a three-step test is proposed (summarised):

  • determine any common matter between the claims;
  • compare the common matter with the “prior art at hand” – does it make a contribution over that prior art and therefore does it comprise any “special technical features”?; and
  • if the common matter does not comprise any special technical features, is there a unifying technical relationship among some of the claims based on remaining features that are not part of the common matter?

In fact, the new Guidelines go so far as to introduce a very welcome new section (F-V, 3.3.1: “Minimum requirements for reasoning lack of unity”) which stipulates that the Examining Division must provide a minimum reasoning outlining their assessment of the claims according to the above steps.

For steps (i) and (ii), this requires the Examiner to identify the common matter, if any, and the reasons why the common matter cannot provide a single general inventive concept. For step (iii), this requires the reasons why there is no technical relationship among the remaining features to be identified, including (for each group) what those remaining technical features are, what problem they solve, and why the problems solved are different.

Applying the above 3-step test to a not-novel independent claim (such as the example above) gives the following result: the independent claim (being common to all dependent claims – step (i)) is not novel over the prior art at hand (and so does not comprise any special technical features – step (ii)). The assessment of unity among dependent claims may then proceed according to step (iii).

When determining whether a feature is a “special technical” one, the new Guidelines caution Examiners against formulating the technical problem (for the inventive step assessment) too generally, leading to the possibly wrong conclusion of a lack of unity.

Step (iii) of the revised test is the clearest stepping-stone to dividing the claims into groups. Indeed, the new Guidelines state that if step (ii) leads to a finding of a lack of unity then the subsequent step (iii) involves identifying groups of potential different inventions. The grouping is done based on the technical problems that are associated with the remaining features of the claims (“remaining” in that they were not considered as part of steps (i) and (ii)), with any claims comprising remaining features that are deemed associated with the same technical problem being assigned into the same group. Whether the remaining features are deemed novel or inventive over the prior art at hand is irrelevant for the grouping.

The purpose of step (iii) is therefore to arrive at different groups of inventions, where each group relates to common subject-matter deemed unitary in view of the prior art at hand.

Do the Changes Clarify the EPO’s Approach?

Whilst some inconsistencies remain, and the possibility for circular argumentation is not eliminated entirely (if a second group of inventions was deemed unified then surely a special technical feature must be present according to Rule 44 EPC…), the new Guidelines are seen as a clarification of the approach taken by the EPO to unity.

Certainly, we consider that they should ensure a clearer and more detailed explanation from the Examiner as to why they have come to the conclusion that there is a lack of unity.


In practice, the clarified Guidelines may give Examiner’s a more structured framework to raise a unity objection amongst dependent claims, particularly if the independent claim(s) are deemed to lack novelty and inventive step. At the moment, we do not see many Examiner’s performing a unity assessment of dependent claims in the event that the independent claim is deemed to lack novelty, so we would certainly hope that the revised Guidelines would not encourage them to do so.

Nevertheless, parts of the new Guidelines could be used to aid the applicant in this scenario. For instance, if features in different claim groups could be argued to be “corresponding” (rather than “the same”) and to provide the same technical effect and/or solution to the same technical problem as the claims in another group, then the new Guidelines support those claims being combined into a single group: if, in the course of grouping, the same special technical feature, which provides a contribution over the prior art, is identified in two groups of inventions, both groups of inventions need to be combined into one single group (from F-V, 3).

It may also be possible to successfully argue that, in alleging lack of unity, the Examiner has formulated the technical problem to be solved (which is associated with the technical effect of the features in question) too narrowly, or too generally, leading to an erroneous conclusion that the claims have nothing in common, and therefore that the technical features are not corresponding. Helpfully, the Guidelines caution against the technical problem being formulated too generally when it could have been formulated more narrowly (F-V, 3).

As mentioned above, even if a lack of unity objection amongst dependent claims can be overcome, the Applicant may face an undesirable consequence of the objection being raised in the first place: if certain claims were deemed not unified and, as a consequence, not searched, then it is possible the Examiner would request their cancellation (for example, under Rule 137(5) EPC). Certainly, we have seen some Examiners make precisely this objection and request, and the Guidelines affirm them doing so: subject-matter that has not been searched will not be examined by the examining division and cannot be prosecuted in an independent claim. If the lack of unity persists also in examination after the claims have been amended, the excision of the non-searched subject-matter from the application may be necessary (F-V, 4.2).

In this instance, it may be worth reminding the Examiner that the wording of Rule 137(5) EPC states that amendments may not relate to unsearched subject-matter which does not combine with the originally claimed invention or group of inventions to form a single general inventive concept. Thus, Rule 137(5) EPC and unity are related, and no claims, even if unsearched, should require cancellation provided it can be argued that they relate to the same inventive concept. Unsearched subject matter can still be pursued in a divisional application, so this situation is not fatal to prosecuting such subject matter, but this is clearly an added expense which is best avoided if at all possible.

A final practice-point that is worth some discussion is that, in this author’s experience, where two dependent claims are both dependent on an earlier claim but where the dependent claims are not dependent on each other (e.g. claim 2 depends on claim 1 and claim 3 only dependents on claim 1, etc.), this can sometimes be seen by the Examiner as a signpost that their subject matter is not combinable (we have seen this, for example, in added matter objections). If the common matter of independent claim 1 is however deemed to lack novelty, then this can lead to a unity objection being raised between the dependent claims, since the Examiner sees their lack of cross-dependency as a signpost that they do not relate to the same embodiment.

Indeed, in a new section (F-V, 3.2.3, “Dependent claims”) this is clarified. Whilst this section affirms that if a higher-ranking claim is not patentable unity must be examined amongst all claims dependent on it, it also states: A dependent claim and the higher-ranking claim on which it depends cannot be grouped into two different groups of inventions.

Therefore, we recommend that multiple dependencies are used in claims wherever this is possible without causing problems in other jurisdictions in which the same specification will be filed.


Some clarification on the EPOs approach to unity was overdue, so in that sense, these new Guidelines are welcome. Whilst the new Guidelines could lead to an increase in unity objections, particularly among dependent claims if an independent claim is deemed to lack novelty, the stipulation (in F-V, 3.3.1) that the Examiner is required to give a more detailed and step-by-step analysis of the unity of the claims is helpful, and should lead to increased certainty not just for the EPO but for the applicant as well.

[1] A further search fee was paid to effectuate this.

This is for general information only and does not constitute legal advice. Should you require advice on this or any other topic then please contact or your usual Haseltine Lake Kempner advisor.