How this guide was made: drafted with AI assistance from the official sources linked below, then passed an automated editorial check before publishing. Our process

Part B of the Horizon Europe proposal template is the technical description that evaluators read and score. It has three numbered sections, each matching one award criterion: 1. Excellence, 2. Impact and 3. Quality and efficiency of implementation. Under the 2026–2027 Work Programme it is capped at 40 pages for Research and Innovation Actions (RIA) and Innovation Actions (IA), and at 25 pages for Coordination and Support Actions (CSA) (General Annexes 2026–2027). Take the template from the submission system for your topic, write each subsection against the scoring wording below, and upload it as a single PDF.

Which option is right for you?

Choose your starting point based on how much experience you have and how much time you can give it.

The reasoning behind each row: beginners need to learn the scoring language first. Regular writers save the most time by checking every paragraph against the evaluation form. Two-stage first stages are judged only on Excellence and Impact. Lump-sum topics follow different page and budget rules. NCPs are national advisory contacts for Horizon Europe applicants; ask yours what services it offers and on what terms.

What is Part B of the Horizon Europe proposal template?

Part B is the narrative and technical part of the proposal. You upload it as a PDF, and it sits alongside the Part A administrative forms you fill in online. According to the Online Manual, Part A holds the administrative forms and Part B is the "Technical Description", uploaded with any annexes in the submission wizard.

In practice the work splits like this. Part A covers participants, budget figures, the abstract, the ethics and security tables and declarations. Part B covers what you will do, why it matters and how you will deliver it. Evaluators score Part B, and they also check it against Part A. If your budget says a partner has 60 person-months and the Part B effort table says 45, someone will notice.

Here is how a proposal moves through the Portal, using the Online Manual's submission steps:

1. Create an EU Login account. Every partner organisation needs a PIC (Participant Identification Code), which you get from the Participant Register.
2. On your topic page, click Start submission to open the submission wizard.
3. Pre-register the draft proposal, then add the participants and contact persons.
4. Fill in the Administrative Forms (Part A).
5. Download the Part B template from inside the submission system. The manual says to use that copy rather than a generic one.
6. Upload Part B as a PDF, together with any annexes the call requires, then click Submit.
7. Wait for the acknowledgement of receipt.

Until the deadline you can submit "as many times as you want". Each new version replaces the previous one. To change a submitted proposal, go to My Proposals > View Submitted > Re-edit proposal. This doesn't work for continuously open calls (Online Manual).

For example, a consortium aiming for a May deadline could upload a complete draft PDF a week early. That checks the upload works and protects you against a failed upload on the last day. You then keep replacing it with improved versions.

How is Part B structured section by section?

Part B has three main sections, each with two or three fixed subsections taken straight from the official template. Most supporting information, such as researcher details, ethics and security, now goes in Part A rather than in Part B annexes.

The page counts below are the template's guideline allocations for RIA/IA under the 2026–2027 version, as summarised by Accelopment and Proposia (as of October 2026; check the template inside your call). Only the overall limit is enforced. The per-section figures are guidance.

Three changes in the 2026–2027 template are worth knowing if you are reusing an older proposal. Section 2.1 no longer asks for the "scale and significance" of your contribution. Tables 3.1i (other cost categories) and 3.1j (in-kind contributions) have been removed. Section 2.3 is now optional (Future Needs; Accelopment). If you copy text from a 2024 proposal, you will be writing about things that are no longer requested and spending pages you no longer have.

Formatting follows the template's own rules: A4 pages, Times New Roman (or the equivalent named in the template) at a minimum of 11 points, and margins of at least 15 mm (Standard Application Form; GrantForge summary). Tables, figures and references sit inside the sections, so they count toward the limit.

To set up your working document:

1. Open the template downloaded from the submission system, not a copy a colleague forwarded.
2. Keep every heading number and title exactly as given. Evaluators look for 1.2, 2.1 and so on by number.
3. Leave the template's instruction boxes in your working draft, and delete them only in the final PDF.
4. Next to each heading, paste the matching sentence from the evaluation form. That sentence is what the reviewer will be scoring.

What goes in the Excellence section and what do evaluators score there?

Excellence (section 1) covers your objectives, your ambition beyond the state of the art and your methodology. Evaluators score how clear and pertinent the objectives are, how sound the methodology is and how credible your claims of going beyond the state of the art are.

The RIA/IA evaluation wording, as quoted by GrantForge from the General Annexes, has two parts. The first is the "clarity and pertinence of the project's objectives, and the extent to which the proposed work is ambitious, and goes beyond the state of the art". The second is the soundness of the methodology, including the underlying concepts, models and assumptions, interdisciplinarity, the gender dimension and open science practices.

1.1 Objectives and ambition

Write objectives that someone else could check at the end of the project. A workable pattern is one sentence per objective, saying what will exist, by when, measured how, and in which work package. State your starting and target Technology Readiness Level (TRL) if the topic mentions TRLs. Then describe the state of the art in a few paragraphs, with citations, and say exactly where your work goes further.

Example (illustrative): "O2: Validate the sensor prototype in two operational water utilities, raising it from TRL 4 to TRL 6 by month 30, measured by detection accuracy ≥ 95% against reference lab samples (WP4, D4.3)." Compare that with "advance water monitoring". An evaluator can score the first sentence.

1.2 Methodology

This is the biggest single block, about 14 guideline pages. According to Proposia, the template splits it into the overall approach (~10 pages), links to other R&I projects (~1 page), interdisciplinarity (~0.5), the social sciences and humanities contribution (~0.5), the gender dimension (~1) and open science (~1). Research data management also belongs here.

How to write it:

1. Start with the concept: the hypothesis or core idea, and the assumptions it depends on.
2. Walk through the approach in the same order as your work packages, so evaluators can cross-reference 3.1.
3. Name the methods, datasets, pilot sites and tools you will use, and say why they are suitable.
4. Give each of the gender, open science and data management subheadings its own short paragraph. If gender is genuinely not relevant, explain why in one or two sentences instead of leaving the heading out.

Pitfalls and fixes

• Objectives stated as activities ("to organise workshops"). Fix: rewrite each one as a measurable result.
• A state of the art with no citations, or more than three years old. Fix: cite recent work, including competing EU projects found on CORDIS.
• Gender dimension mixed up with gender balance in the team. Fix: gender dimension means sex and gender analysis in the research content itself. Team balance is a separate question.

What goes in the Impact section and how is it scored?

Impact (section 2) explains how your results lead to the outcomes and impacts named in the work programme, and how you will disseminate, exploit and communicate them. Evaluators score how credible your pathways are and how suitable your measures to maximise impact are.

Since the 2026–2027 template, section 2.1 asks for the pathway (who benefits, how and why) and no longer asks you to quantify "scale and significance" (Future Needs). That doesn't make numbers useless. A credible pathway often still uses a few well-sourced figures. The difference is that you no longer need to fill a dedicated paragraph with market-size estimates.

For Innovation Actions, Impact carries a weight of 1.5 when proposals are ranked, though not when thresholds are checked (General Annexes, as summarised by GrantForge). In an IA, an extra half-point on Impact is worth more in the ranking than an extra half-point anywhere else.

2.1 Project's pathways towards impact (~3 pages)

1. Copy the "Expected outcomes" bullets from your topic text and the "expected impacts" from its Destination.
2. For each outcome, name the result that produces it, the user who will adopt it, and when.
3. Name the barriers, such as regulation, cost, skills or public acceptance, and say what you will do about each.

2.2 Measures to maximise impact (~3 pages)

Cover the three activities separately:

• Dissemination: sharing results with people who can use them, for example open-access papers, datasets and standards bodies.
• Exploitation: who will use which result after the project, through licensing, products, policy uptake or follow-up projects.
• Communication: reaching the wider public.

Add your intellectual property strategy, meaning who owns which result and how it will be protected.

2.3 Summary (optional)

If you use it, the summary is a one-page canvas linking needs, target groups, outcomes, impacts and your dissemination, exploitation and communication measures. Since the 2026–2027 template it is optional (Accelopment). If you are short of pages, cutting it is reasonable. If an evaluator needs a quick view of a complex pathway, it can help.

Pitfalls and fixes

• Generic dissemination lists ("website, social media, conferences"). Fix: name the target groups, the channels and a measurable indicator, for example "3 policy briefs to DG ENV and national water regulators by month 24" (illustrative).
• Impact claims for outcomes the topic never asked for. Fix: tie every claim to the wording of your topic.
• Exploitation described only for the coordinator. Fix: give each partner a line saying what it will do with which result.

What goes in the Implementation section and what do evaluators check?

Implementation (section 3) covers the work plan, the tables of work packages, deliverables, milestones, risks and effort, and the consortium's capacity. Evaluators check whether the plan can be delivered, whether risks are managed and whether the partners complement one another.

The evaluation wording covers the quality and effectiveness of the work plan, the assessment of risks, whether the effort assigned to each work package is appropriate, and the capacity and expertise of the participants (GrantForge).

3.1 Work plan and resources (~12 pages, 17 for lump sum)

Start with a short overview, a Gantt chart and a PERT diagram showing how the work packages connect. Then fill in the tables. The table structure below is described by Proposia and Accelopment:

How to build it:

1. Agree the work package list and person-months with all partners before anyone starts writing prose.
2. Fill in 3.1f first, then check it against the personnel figures each partner enters in the Part A budget.
3. Number deliverables Dx.y to match their work package, and give every objective in 1.1 at least one deliverable or milestone.
4. Write 3.1e risks that are specific to this project, such as "pilot utility withdraws". "Delays" is too generic. Pair each risk with a named fallback.

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

Explain why each partner is in the consortium, what it brings that the others don't, and how the consortium covers the whole value chain the topic implies, for example end users for an IA. Detailed researcher information goes into Part A, so 3.2 should be about how the partners complement each other, not a list of CVs.

Pitfalls and fixes

• A WP leader with 2 person-months in their own WP. Fix: give WP leaders effort that matches the job of leading.
• More milestones than decision points. Fix: about 4 to 8 milestones, each with a "means of verification", is easier for evaluators to trust than 20 (an editorial suggestion, not a rule).
• Using the deleted tables 3.1i or 3.1j from an old template. Fix: start again from the current template.

How do page limits and formatting rules differ by call type (RIA vs IA vs CSA)?

Page limits depend on the type of action and on whether the topic uses lump-sum funding. Anything over the limit is hidden from evaluators after the deadline.

According to the General Annexes 2026–2027, "excess pages will be automatically made invisible". Limits as of October 2026 (check your topic page, because individual topics can set different ones):

The 2023–2025 figures come from Accelopment.

Scoring is the same for every action type. Each criterion gets 0–5, with half points allowed. The default threshold is 3 per criterion and 10 overall. In the first stage of a two-stage call only Excellence and Impact are evaluated, and the threshold for each is 4 (HE evaluation briefing; Online Manual PDF). Three scores of exactly 3 add up to 9, which is below the overall threshold.

The CSA template keeps the same three sections, but section 1 is aimed at coordination and support measures rather than research beyond the state of the art (GrantForge). Don't use the RIA template for a CSA topic. Download the variant your topic links to.

Steps to confirm your limit:

1. Open your topic page on the Portal and scroll to Topic conditions and documents.
2. Read the type of action (for example "HORIZON-RIA" or "HORIZON-CSA") and check whether lump-sum funding applies.
3. Check the topic's specific conditions for a different page limit.
4. Count your pages in the final PDF, including the title page and the participant list.

How do you fill out the mandatory annexes (CVs, ethics, security)?

In Horizon Europe most "annex" content goes into the Part A forms. The ethics self-assessment and the researcher information are entered online, and topic-specific annexes are uploaded only if the call asks for them.

Ethics is handled in two steps in the administrative forms (section 4). First you answer the Ethics Issues Table. Then, for every "Yes", you explain in the Ethics Self-Assessment how you will deal with the issue (UZH Ethics factsheet, April 2025). The Commission's guide "How to complete your ethics self-assessment" is the reference document for this, and it is linked from the Portal's reference documents.

Researcher details are entered in a researchers table in Part A. That table lists only the researchers involved, not administrative staff (RVO template v4.0). Part A also asks each participant for relevant publications, previous projects and significant infrastructure. This is why section 3.2 of Part B should not repeat CVs.

Steps:

1. Go through the Ethics Issues Table with every partner, not just the coordinator. Human participants, personal data, animals, non-EU countries and dual use are the questions most often answered "Yes".
2. For each "Yes", write what you will do, who approves it (for example the local ethics committee) and when. Refer to the relevant work package.
3. Answer the security questions in the same administrative forms. If your topic involves classified information or security-sensitive results, read the topic conditions and the General Annexes for any extra documents required.
4. For lump-sum topics, upload the detailed budget table the call requires (GrantForge).
5. Check that the personal data issues ticked in Part A also appear in your data management text in 1.2.

Pitfalls: answering "No" to personal data while the methodology describes interviews; writing ethics text in Part B where nobody reviewing ethics will look for it; and leaving partner-specific issues, such as a non-EU partner's data transfers, out of the self-assessment.

What are the most common mistakes that lower Part B scores?

Most lost points come from going over page limits, vague impact pathways, Part A and Part B figures that don't match, and subsections the topic requires but the proposal leaves out. All of these can be fixed before submission.

The evaluation summary report (ESR) you receive gives comments under each criterion, using the 0–5 descriptors from "Fails" to "Excellent". Score 3 is described as "addresses the criterion well, but a number of shortcomings are present" (GrantForge, quoting the scoring scale). Fixing the mistakes below is how you move from a 3 to a 4.

Example (illustrative): a proposal whose 1.1 promises a TRL 6 demonstration, but whose 3.1c lists no demonstration deliverable, invites a comment such as "the link between objectives and work plan is not sufficiently clear". The fix takes one line in the deliverables table.

A final check before you click Submit:

1. Every heading from the template is present and in the original order.
2. Every objective links to a WP, a deliverable and a KPI.
3. Every expected outcome of the topic is mentioned by name in 2.1.
4. Person-months and costs match Part A.
5. The PDF opens, is searchable, and is within the page limit.

Next steps

You can do the first round of work today without spending anything.

DIY actions for today:

1. Open your topic page, note the action type and whether it is lump sum, and write down your page limit (40, 45, 25 or 28).
2. Download the Part B template through Start submission in the submission system, and save the evaluation form next to it.
3. Draft table 3.1a (the work package list) and table 3.1f (effort) with your partners before writing any prose.

If you would rather not do it alone: your National Contact Point gives advice and often reviews drafts (as of October 2026; services and terms vary by country, so ask yours what it offers). This suits first-time applicants and SMEs. It is less suited to anyone who needs someone to write the text for them, because NCPs advise rather than write. SMEs can also contact the Enterprise Europe Network, which suits companies looking for partners or innovation support; check the service terms with your local EEN office. Its depth on technical Part B review varies. Private grant consultants will write or edit Part B for a fee. Fee structures vary (day rates, fixed fees, success fees), so ask for a written quote and check their track record on CORDIS. Consultants suit consortia short on time. They are a poor fit if your budget is tight or if the scientific content isn't settled yet, because nobody can write a strong methodology the team hasn't agreed on.

FAQ

Where can I download the official Horizon Europe Part B template?

Get it from inside the Funding & Tenders Portal submission system, after you click Start submission on your topic page. The Online Manual says to use that copy (Online Manual). You can read the general RIA/IA standard application form beforehand, but always submit using your topic's version.

Is Part B submitted as a PDF or filled online?

Part B is uploaded as a PDF in the submission wizard. Part A is filled in online as web forms (Online Manual). You can upload a new PDF as often as you like until the deadline, and the latest version replaces the earlier ones.

What is the page limit for a Horizon Europe RIA proposal?

Under the 2026–2027 Work Programme, a full RIA proposal is limited to 40 pages, or 45 for lump-sum topics. A first-stage proposal in a two-stage call is limited to 10 pages (General Annexes). Individual topics can set a different limit, so check your topic page.

Do I need a different Part B template for different calls?

Yes. RIA/IA, CSA, lump-sum, two-stage and programme-specific actions such as the EIC and MSCA each have their own forms. The Commission publishes them separately, for example the MSCA-PF form and the EIC Accelerator form. Use the one linked from your topic.

What is the difference between Part A and Part B of a Horizon Europe proposal?

Part A is the administrative part: participants, budget, abstract, ethics and security tables, and declarations, all completed online. Part B is the technical description that evaluators score against Excellence, Impact and Implementation. The two must agree on figures such as person-months and costs.

Sources

• EN Annex XV, Horizon Europe Work Programme 2026–2027, General Annexes
• Standard Application Form (HE RIA, IA), European Commission
• Horizon Europe Evaluation Form (HE RIA and IA)
• Horizon Europe proposal evaluation: standard briefing slides for experts
• Online Manual: Electronic proposal submission
• Online Manual: Submit a proposal (Electronic Submission System)
• EU Funding & Tenders Online Manual, v1.1 (PDF copy)
• RVO: Horizon Europe Standard Application Form v4.0 (annotated)
• UZH: Ethics Assessment in Horizon Europe factsheet (April 2025)
• Accelopment: Horizon Europe 2026–2027 template changes
• Future Needs: New template for Horizon Europe proposals 2026–2027
• Proposia: Horizon Europe Part B template, every section
• Proposia: Work packages, Gantt and PERT
• GrantForge: The Part B template, section by section
• GrantForge: How Horizon Europe proposals are scored
• Horizon Europe NCP Portal
• Enterprise Europe Network
• CORDIS

Originally published at Call Window EU Grants.