Horizon Europe Two-Stage Calls: How to Write the 10-Page, Blind, Stage-1 Proposal

A two-stage call asks for a short proposal first and a full proposal only from those who pass. For applicants that is good news, because a failed stage 1 costs ten pages rather than forty. The catch is that the ten pages are scored against a higher bar, on only two criteria, and in most cases without the evaluators knowing who wrote them. This page covers the numbers, the blind-evaluation rules in the Commission’s own words, a page budget built on the actual template sections, and a checklist to run on the PDF before you upload it.

What a two-stage call is, and why the Commission uses them

The general introduction to the 2026-2027 work programme counts 41 call topics with two-stage evaluation. It explains the purpose in two sentences: applicants submit a shorter summary proposal and a full proposal only if successful, and many of the short proposals are evaluated blindly, through anonymised applications, to avoid potential bias. Shortening the first-stage submission and evaluation also shortens the time to grant. The work programme’s other structural changes, including lump sums and the new page limits, are covered in the Horizon Europe 2026 guide.

Stage 1 and stage 2 compared

Two consequences of the higher bar are worth stating. A 3.5 on Excellence, a passing score in any single-stage call, fails stage 1 outright. And because the combined threshold is sized to admit two to three times the budget, stage 2 is still a real competition, not a formality. How scores, thresholds and evaluation summary reports work in general is explained in how Horizon Europe proposals are scored.

Stage 2: what may change and what may not

The full proposal must build on the short one. NCP Flanders, quoting Commission guidance, gives acceptable changes as the absence of one or two of the partners named at stage 1, or a change of coordinator. It gives unacceptable changes as a completely changed conceptual basis or methodology, significantly altered objectives, or an expected impact that no longer corresponds to the stage-1 text. The evaluator briefing adds a rule that works in your favour: experts at stage 2 must avoid inconsistencies with the stage-1 assessment and justify any difference of opinion. In practice, write stage 1 as the first ten pages of the full proposal, not as a pitch you intend to rewrite.

Blind evaluation: the rules in the Commission’s words

Blind evaluation is the default method for first-stage proposals of two-stage calls, but exceptions are possible, and it applies only where the topic conditions say so. Where it applies, the blind template’s guidance is unambiguous: proposals may not contain any direct or indirect reference to the organisations or persons participating. Applicants must not disclose organisation names, acronyms, logos or names of personnel in the abstract or Part B, and must not be identifiable indirectly, for instance through links to web pages or references to their role and previous experience. Identity belongs only in the Part A web forms. The system hides those from evaluators, with one exception that catches applicants every round: the abstract is transmitted.

A direct identification makes the proposal inadmissible. Admissibility checks, including anonymity, are usually done by Commission or agency staff; indirect identifications are analysed case by case and may also lead to inadmissibility. Evaluators, for their part, are told not to guess or investigate who is behind a proposal, and that listing a past project does not imply the applicants took part in it.

Citing your own work

You may cite your own publications, provided nothing emphasises that the proposers wrote them. The template gives the canonical pair.

Sentences the Commission lists as inadmissible

The template and the evaluator briefing both reproduce the same list of real-world failures. Every one is an ordinary sentence in a non-blind proposal.

• “Most of project’s participants have been involved in the previous H2020 project, NANOCOM...”
• “This task in WP3 will be based on outputs generated by some participants of the consortium”, with a footnote link to a web page or video that identifies them
• “The consortium includes the largest research institute in France”
• “Partner 3 is the leading company in Spain for wind turbine installation”
• “Our current research expands our previous findings described in a recently published article (Wiliam et al, 2022)”
• “The consortium consists of leaders in the high tech industry, including the biggest in terms of capital constructor of micro chips”
• “The coordinator organisation was the one who first introduced the concept of m-RNA in vaccines”
• “The consortium consists of 2 research centres (including an international one based in Geneva) and the oldest university in Belgium”

The briefing also addresses two grey areas. Past projects can be named, but you cannot say the proposal continues them or that their partners overlap with yours; if you want to present an activity as your own, do not name it and describe it so it cannot be identified. A place where the work happens can be named, as long as the proposal does not say any applicant is from there; otherwise describe the location by its characteristics.

A page budget for ten pages

The first-stage template has exactly three sections, not the six of the full proposal: 1.1 Objectives and ambition, 1.2 Methodology, and 2.1 Project’s pathways towards impact. The template calls its page counts indicative and gives none for stage 1, so the split below is a working allocation, weighted toward the sections evaluators most often mark down.

Note what is missing. The first-stage Excellence criterion looks at the clarity, pertinence and ambition of the objectives and the soundness of the overall methodology; the Impact criterion looks only at the credibility of the pathways. There is no implementation, no work plan and no consortium section, and critical risks of project management belong to the full proposal. To check whether your text fits, the text-to-space estimator converts word counts into pages at the required 11-point font.

Writing for two criteria and no names

Objectives that reach a 4

A 4 means the criterion is addressed very well with a small number of shortcomings. For objectives, that means each one is measurable, realistically achievable and visibly tied to a named expected outcome of the topic, so the evaluator can tick the topic text line by line. A short table of objective, target, and the expected outcome it serves does that in a third of a page.

Methodology without a consortium

The blind template asks for descriptions to be given in general terms to preserve anonymity. Describe capabilities, not owners: a certified manufacturing facility, a cohort of a stated size, access to a national registry, a validated model. The quantity and quality of the asset carry the argument; who holds it is for stage 2. Counting and typing the consortium, for example nine organisations including SMEs and hospitals, is not on the Commission’s list of identifying statements, but superlatives about any of them are.

An impact pathway that is specific

The stage-1 template asks for the unique contribution of your results to the topic’s outcomes and the destination’s impacts, with quantified estimates where meaningful, target groups broken into specific segments, and the barriers outside the project. Use the destination’s own wording for impacts so the evaluator does not have to translate. The difference between results, outcomes and impacts is set out in the Part B template walkthrough.

Before submitting, check that the ten pages address every expected outcome in the topic text. Proposia’s Alignment Review does that comparison against the call.

The anonymisation checklist

Run this on the final PDF and on the abstract as it appears in Part A, not on the Word file. Most leaks are in places nobody rereads: figure labels, file properties and footnotes.

DIRECT IDENTIFICATION (inadmissible)
[ ] No organisation names in text, tables, figures, captions, headers, footers or file name
[ ] No organisation or project acronyms that point to a participant, including your previous projects
[ ] No logos, letterheads or institutional slide templates in figures
[ ] No personnel names, ORCID iDs, email addresses or signatures
[ ] PDF properties (Author, Title, Company) cleared; comments and tracked changes removed

INDIRECT IDENTIFICATION (case by case, may be inadmissible)
[ ] No "developed by a partner", "our previous findings", "our lab" next to a citation
[ ] No statement that the proposal continues a named project or shares its partners
[ ] No superlatives that single out an organisation: "largest", "leading", "oldest", "first to"
[ ] No links or footnotes to web pages, videos or repositories that name participants
[ ] Locations described by characteristics if naming them would reveal where a partner is
[ ] Assets described by function and scale, not by a unique proper name

ALLOWED
[ ] Citations to your own work formatted like any other citation
[ ] Names of past projects, without claiming them as yours
[ ] A description of the consortium by number and type of organisation, without superlatives

FILES AND FORMS
[ ] Abstract in Part A anonymised: it is transmitted to evaluators
[ ] Part B built on the HE 1st stage RIA and IA BLIND template for this topic
[ ] Topic conditions confirm blind evaluation (keep a screenshot)
[ ] Final read by someone outside the consortium, looking only for identity leaks

Timeline: from stage-1 deadline to grant

The rhythm is slow and worth planning around: roughly sixteen months from the stage-1 deadline to a signed grant agreement, with almost a year of it spent waiting.

One consequence for planning: if a first-stage proposal fails, the generalised feedback is thin and the full evaluation summary report only arrives with the stage-2 result, so there is little to work with before the next call. For teams resubmitting after a stage-2 rejection, the Resubmission Assistant maps ESR comments to the sections they concern. Consortium changes between stages are easier when the partners were chosen well; see finding Horizon Europe partners.