The Horizon Europe Part B Template, Section by Section (2026-27)

Part B is the document the experts actually read. Part A is the set of web forms on the Funding & Tenders Portal: administrative data, the budget table, the ethics and security self-assessment. Part B is the PDF you upload, and its three numbered sections map one-to-one onto the three award criteria of Excellence, Impact and Quality and efficiency of the implementation. Each section is scored from 0 to 5, so the structure of the template is also the structure of your score. If you want to see how those scores are produced, read the companion page on how Horizon Europe proposals are scored and what the Evaluation Summary Report contains.

This page walks the current template, version 5.1 of 22 January 2026, a formatting update to version 5.0 of November 2025, which introduced the 2026-27 changes, heading by heading. Where the template gives an indicative page count, it is quoted. Where the evaluator briefing tells experts what to look for, that is quoted too, because the two documents are written to mirror each other and the fastest way to write a fundable section is to answer the question the expert has been told to ask.

What changed for 2026-27

Three things matter if you are adapting a Part B written under the 2023-25 work programmes. First, the page limit for RIA and IA proposals is now 40 pages, or 45 for topics funded as lump sums; the earlier 45/50 allowance is gone. Second, the impact criterion was simplified: from the 2026-27 work programme onwards the scale and significance of your contribution to the expected outcomes is no longer an aspect experts score, only the credibility of your pathway. Third, two requirements were dropped from the template altogether, the Do No Significant Harm statement and the AI robustness paragraph; possible environmental consequences are still considered under Excellence and Impact where relevant, and AI-related ethics questions moved to the Part A ethics table. The structured tags of earlier versions were also removed, so the headings below are the only structure you need to reproduce.

One thing did not change: the template still says proposals are evaluated as submitted, not on their potential, and that there is no possibility of significant changes to content, budget or consortium during grant preparation. Every gap you leave is a gap the experts will score.

The page budget

The limit covers the title, the list of participants and sections 1, 2 and 3 together, and every table, figure and reference list inside those sections counts. The template calls its own per-section numbers indicative, but they add up to the limit almost exactly, which makes them a useful allocation for a first draft.

The formatting rules are enforced as strictly as the page count. Body text, including text inside tables, must be at least 11 points in Times New Roman or its Apple and Linux equivalents, single spaced, on A4 with 15 mm margins on every side. A different body font is tolerated only if it is legible and does not shorten the document compared with the reference font. Uploading an over-long file before the deadline triggers a warning; after the deadline the extra pages are simply made invisible to the experts. The template adds a line that applicants often skip past: do not treat the page limit as a target, because experts rarely view unnecessarily long proposals in a positive light.

Section 1: Excellence

Experts assess two aspects here: the clarity and pertinence of the objectives together with how far the work goes beyond the state of the art, and the soundness of the methodology, which explicitly includes the gender dimension in research content and the quality of open science practices. Both are judged only to the extent that the work is within the scope of the topic, so the scope check comes first.

1.1 Objectives and ambition (about 4 pages)

The template asks three questions of your objectives: are they pertinent to the topic, are they measurable and verifiable, and are they realistically achievable. The evaluator briefing repeats them word for word. The ESR analysis by the Horizon Academy NCP network found the same three phrases in the comments of low-scoring proposals, usually as an absence: objectives not linked to the call’s expected outcomes, no key performance indicators, or KPIs stated without any explanation of how they would be calculated. A short table of objectives, each with a measurable target and the work package that delivers it, answers all three questions in half a page.

The second half of 1.1 is the advance beyond the state of the art. Name the ground-breaking element if there is one, refer to products already on the market where relevant, and mention any patent or publication search. Then position the work on the idea-to-application spectrum and give a Technology Readiness Level at start and end. The template is explicit that expectations differ by positioning: a low-TRL research action is not expected to show the same advance as an innovation action at TRL 7.

1.2 Methodology (about 14 pages)

This is the largest block in the proposal and the template splits it into six required elements, each with its own indicative length.

• Overall methodology, about 10 pages. Concepts, models and assumptions, how they deliver the objectives, and the important challenges you foresee and how you will overcome them. Written as a narrative; the tasks belong in section 3.
• Links to other research and innovation activities, about 1 page. National or international projects whose results feed into yours, and how the link is established. If your proposal is being evaluated blind, name the projects without implying you were a partner in them.
• Interdisciplinarity, about half a page. How methods from different disciplines are integrated, or a justification that an interdisciplinary approach is unnecessary.
• Social sciences and humanities, about half a page. Required where the topic flags SSH integration; the Horizon Academy analysis recommends a short justification even where it does not.
• Gender dimension, about 1 page. Sex and gender analysis in the research content, or a justification of why it is not relevant. Mandatory unless the topic says otherwise, and unrelated to team gender balance.
• Open science practices, about 1 page. Open access to publications and FAIR data management are mandatory in every call and lower the score when not addressed; early sharing, reproducibility measures, open peer review and citizen engagement are recommended and score positively when present. Outreach belongs under Impact, not here.

Two follow-on obligations are worth stating in the text because experts check for them: a data management plan is due as a deliverable by month 6 and revised toward the end of the project, and if generative AI tools were used to prepare the proposal the template expects the content to have been reviewed and validated by the applicants for accuracy and compliance.

Section 2: Impact

Two aspects are scored: the credibility of the pathways to the expected outcomes and impacts specified in the work programme, and the suitability and quality of the measures to maximise them. The vocabulary is defined in the template and it pays to use it precisely.

2.1 Project’s pathways towards impact (about 3 pages)

Write a narrative with two parts. Part (a) is your unique contribution to the outcomes listed in the topic and to the longer-term impacts of the destination, with quantified estimates where meaningful, and with target groups broken into specific segments rather than the generic groups named in the work programme. Part (b) is the requirements and barriers beyond your control, such as other R&I work, regulation, target markets and user behaviour, how they may evolve, and the mitigating measures inside or outside the project. The template draws a line here that many proposals blur: critical risks to delivering the project belong in section 3, not in 2.1.

Since the 2026-27 work programme, experts no longer score scale and significance, so a paragraph of inflated market-size claims earns nothing. Credibility does: the French Health NCP’s analysis of 189 Evaluation Summary Reports found the pathway to impact was the single most commented aspect, with the recurring negative comments being that it was insufficiently described, not quantified, not scalable, or silent on barriers.

2.2 Measures to maximise impact (about 3 pages)

This section holds the first version of your plan for dissemination and exploitation including communication activities. Unless the topic says otherwise, that plan is an admissibility condition, and a detailed version becomes a mandatory deliverable within six months of signature. The template asks for the planned measures, the target group each one addresses, and a justification of why that measure suits that group. Communication must be planned strategically from the outset with main messages, tools and channels per target group. Innovation actions must describe a plausible path to commercialisation, and exploitation expected mainly in non-associated third countries must be shown to be in the Union’s interest. Close with the intellectual property strategy: foreseen protection measures and how they support exploitation. If funded, you will need a consortium agreement to manage ownership and access rights, and the final periodic report must list the owners of the results.

Section 3: Quality and efficiency of the implementation

The two aspects scored are the quality and effectiveness of the work plan, risk assessment and the appropriateness of the effort and resources, and the capacity and role of each participant together with the complementarity of the consortium as a whole.

3.1 Work plan and resources (about 12 pages, 17 for lump sum, including tables)

The template lists what this section must contain: a brief presentation of the overall work plan structure, a timing chart such as a Gantt, a chart of how the components inter-relate such as a PERT, and the detailed work description in eight numbered tables. It also gives advice that experts treat as expectation: include a distinct project-management work package, give visible tasks or work packages to data management, dissemination and exploitation, and communication, keep the number of work packages proportionate to the scale of the project, quantify every task, and make sure the person-months here match the budget table in Part A. For a fully worked example of a six-work-package plan with these tables filled in, see the work package, Gantt and PERT worked example in this series, and use the Gantt chart creator to draw the timing chart.

Three details in the table key catch out first-time coordinators. Deliverables are numbered by work package and order of delivery, so deliverable 4.2 is the second deliverable of WP4, and the list should contain only what is essential for monitoring, because every deliverable becomes a contractual obligation. The data management plan and the dissemination and exploitation plan must both appear as distinct deliverables within the first six months. And each milestone needs a means of verification that a project officer can check, such as a prototype running, software validated by a user group, or a survey completed with data quality confirmed.

3.2 Capacity of participants and consortium as a whole (about 3 pages)

Do not repeat the participant descriptions from Part A. Instead show how the consortium matches the objectives: the disciplinary and interdisciplinary knowledge it brings, including SSH, open science and gender expertise where appropriate; access to critical infrastructure; how members complement one another along the value chain; and the valid role and adequate resources of each partner. Where industrial partners are present, tie their involvement to the exploitation measures in 2.2. If any participant requesting funding is from a country not automatically eligible, this is where you justify why it is essential. The Horizon Academy ESR analysis notes that evaluators comment negatively on unexplained imbalances in effort and on partners with undefined roles, and positively on scientific advisory boards and on consortia with a track record of previous EU projects together. If you are still assembling the consortium, the consortium survival guide covers due diligence and the DESCA agreement.

Variants: CSA and two-stage calls

Coordination and Support Actions use a separate template with the same three sections but a 25-page limit (28 for lump sum), and the Excellence criterion is reworded to the quality of the coordination and support measures rather than an advance beyond the state of the art. Even so, the Horizon Academy analysis found evaluator comments about state of the art in several CSA reports, so a short paragraph does no harm.

First-stage proposals in two-stage calls use the 10-page first-stage template (also version 5.0, 5 November 2025), which contains only sections 1 and 2, because only Excellence and Impact are evaluated at stage 1, each with a threshold of 4 out of 5. From the 2026-27 work programme these first-stage proposals are evaluated blind: Part A identification data is hidden automatically, but it is up to you to remove every direct or indirect identification of the applicants from Part B and the abstract, and a direct identification makes the proposal inadmissible. The rules for citing your own publications and past projects without revealing yourself are covered in the two-stage and blind calls guide.

Pre-upload compliance checklist

Run this against the PDF you are about to upload, not against the Word file. Every item is taken from a rule in the template or from a recurring negative comment in the two ESR analyses cited below.