Horizon Europe, the European Union’s research and innovation framework programme (2021–2027), makes research data management (RDM) a mandatory Open Science practice for any project that generates or reuses digital research data. The centerpiece of that requirement is a Data Management Plan (DMP) — but the Horizon Europe DMP is not a generic document. It follows a specific European Commission template built around the FAIR principles (Findable, Accessible, Interoperable, Reusable), with its own deadlines, section headings, and reviewer expectations that differ in structure from the NIH- or NSF-style plans US-funded researchers may already be familiar with. This guide covers what the Horizon Europe DMP requires, when it’s due, what the FAIR sections actually ask for, and how it differs from US funder templates.
Is a Data Management Plan mandatory under Horizon Europe?
Yes, for any project that generates and/or reuses digital research data. Research data management is one of the mandatory Open Science practices embedded in the Horizon Europe Model Grant Agreement, and beneficiaries are expected to manage the data they produce according to the “FAIR and open by default” principle, applied under the qualifier “as open as possible, as closed as necessary” (see below). A DMP-like statement, typically brief, is expected at proposal stage in the relevant application section; the full, formal DMP is a mandatory project deliverable once funded.
Timeline: proposal-stage summary vs. the month-6 deliverable
Horizon Europe’s DMP requirement has two distinct moments:
- At proposal stage, applicants address data management in the relevant part of the standard application form — a short, high-level statement of intent rather than the full plan. A formal DMP document is not required to submit a proposal.
- Within the first six months of the project (month 6 from grant signature), a full, detailed DMP must be submitted as a formal deliverable, using the European Commission’s official DMP template (available from the EU Funding & Tenders Portal).
The DMP is treated as a living document: it must be kept updated as data-related decisions change over the project’s lifetime (choice of repository, access conditions, data types generated, and so on). For projects running longer than 12 months, at least one further updated version must be submitted as a subsequent deliverable, typically aligned with reporting periods or major project milestones.
The template structure: what the FAIR Data section actually asks
The official Horizon Europe DMP template is organized around headings that map directly onto the FAIR principles rather than the more open-ended narrative format common in NIH/NSF-style plans. The core sections are:
- Data summary — the purpose of data collection/generation, its relation to project objectives, the types and formats of data produced, whether existing data is being reused, the origin of the data, and its expected scale (volume).
- FAIR data, broken into its own sub-headings:
- Making data findable — discoverability conventions, naming conventions, versioning, and metadata provision, including whether metadata will be searchable via a catalogue or a specific search engine, and what identifiers (e.g. DOIs) will be assigned to make data uniquely and persistently identifiable.
- Making data accessible — which data will be made openly available and which will be restricted (with justification), what methods/software are needed to access it, and where it will be deposited (naming the repository).
- Making data interoperable — the vocabularies, standards, and methodologies used so data can be exchanged and reused across systems and disciplines.
- Increase data re-use — licensing terms (e.g. Creative Commons), the point at which data will be made available for reuse, data quality assurance processes, and how long data will remain reusable after project end.
- Other research outputs — non-data outputs (e.g. software, protocols, physical samples) that also need management.
- Allocation of resources — the costs of making data FAIR, who is responsible for data management within the consortium, and how long-term preservation costs will be covered after the project ends.
- Data security — how data will be stored and secured during and after the project, including relevant provisions for sensitive data.
- Ethical aspects — any ethics- or legal-related issues that affect data management, typically cross-referenced to the project’s ethics deliverables.
- Other issues — reference to any other national/funder/institutional/sectoral procedures for data management the project follows.
This checklist-style, sub-headed structure is the single biggest practical difference from a typical Data Management Plan (DMP) written for a US funder: Horizon Europe reviewers and project officers expect each FAIR sub-heading addressed explicitly, rather than folded into general narrative prose.
“As open as possible, as closed as necessary”
Horizon Europe’s default expectation is open access to research data (commonly under CC BY or CC0 licensing terms), but this default is explicitly qualified: beneficiaries may restrict access where opening the data would conflict with legitimate business interests (e.g. IP commercialization plans), data protection obligations, confidentiality, security concerns, or other legal constraints. Where data is kept closed or access-restricted, the DMP must document the justification rather than simply omit the data from consideration. This “open by default, closed where justified” framing is a core Horizon Europe Open Science principle, distinct from data being FAIR — data can be FAIR (findable, accessible under stated conditions, interoperable, reusable) while still being access-restricted, since FAIR does not require full openness, only that access conditions be clear and, where relevant, machine-actionable.
Deposit and the European Open Science Cloud (EOSC)
Where a call’s conditions specify it, beneficiaries must deposit research data in a repository federated with the European Open Science Cloud (EOSC), the EU’s initiative to build a federated, trusted environment for storing, sharing, and reusing research data across disciplines and member states. Choosing a repository that meets recognized trustworthiness criteria (for example, CoreTrustSeal certification) is good practice generally, and is often the practical route to meeting Horizon Europe’s data-accessibility and long-term-preservation expectations even outside a strict EOSC-federation mandate.
How this differs from NIH/NSF-style DMPs
Researchers and research administrators moving between US and EU funding are often surprised by how differently the DMP requirement is structured, even though the underlying goal — making funded research data well-managed and appropriately shareable — is the same. Some of the practical differences:
- Template shape: Horizon Europe supplies a specific, sectioned template built around the four FAIR sub-principles as distinct headings; NIH’s and NSF’s requirements (see NIH vs. NSF Data Management Plans) are typically narrative policy statements answering a shorter list of prescribed questions, without a FAIR-labelled section structure.
- Timing: Horizon Europe explicitly splits a brief proposal-stage statement from a full month-6 post-award deliverable; NIH’s Data Management and Sharing Policy requires the full plan at proposal/just-in-time stage, not as a later deliverable.
- Explicit FAIR framing: Horizon Europe’s template names FAIR directly and structures the document around it; US funder guidance references data sharing, reuse, and preservation expectations that overlap substantially with FAIR in substance but do not require the plan to be organized under FAIR headings.
- Update obligation: Horizon Europe requires the DMP to be resubmitted as an updated deliverable at least once for longer projects; this formal “DMP as a living, re-deliverable document” requirement is more explicit than typical US funder practice, where updates are usually expected in principle but not required as a separate reportable milestone.
- Infrastructure pointer: Horizon Europe’s guidance points toward EU-specific infrastructure (EOSC-federated repositories) where call conditions require it; NIH/NSF guidance instead points toward domain-appropriate repositories generally, without a single named federated infrastructure.
For background on making a dataset meet the FAIR principles regardless of which funder’s template applies, see How to Make Your Dataset FAIR: A Step-by-Step Checklist. For how FAIR relates to, and differs from, the CARE Principles for Indigenous Data Governance, see FAIR vs CARE.
Frequently asked questions
Does every Horizon Europe project need a Data Management Plan?
Every project that generates and/or reuses digital research data needs one. Research data management is a mandatory Open Science practice under Horizon Europe; the DMP is the deliverable that documents how a given project meets that obligation.
When is the Horizon Europe DMP due?
A brief statement of intent is expected at proposal stage; the full DMP, using the EC’s official template, is due as a formal deliverable within the first six months (month 6) of the project, and at least one updated version is due later for projects running longer than 12 months.
Does a Horizon Europe DMP mean all project data must be made openly available?
No. The default expectation is openness (“as open as possible, as closed as necessary”), but access can be restricted where justified by legitimate business interests, data protection, confidentiality, security, or other legal constraints — provided the DMP documents the justification.
Where do I get the official Horizon Europe DMP template?
The European Commission publishes the official template on the EU Funding & Tenders Portal; consortia commonly draft their plans using machine-actionable DMP tools aligned to the Horizon Europe template, such as OpenAIRE’s ARGOS or DMPonline, though using a specific tool is not itself an EC mandate.
How is this different from a NIH or NSF Data Management Plan?
Structurally, the biggest difference is that Horizon Europe’s template is explicitly organized around the FAIR principles as named sub-headings and splits the requirement into a proposal-stage statement plus a later month-6 deliverable, whereas NIH and NSF plans are narrative documents submitted at proposal stage answering a shorter set of prescribed questions. See NIH vs. NSF Data Management Plans for the US-side comparison.
