Horizon Europe Eligible Costs Explained
A Horizon Europe budget can look credible at proposal stage and still become a problem once the grant agreement is signed. The distinction matters because Horizon Europe eligible costs are not simply expenses a consortium expects to incur. They are costs that meet the rules of the applicable Model Grant Agreement, the action conditions and the beneficiary’s own accounting practices, and that can be evidenced later.
That has two consequences. First, the budget table is part of the proposal’s implementation case: evaluators use it to judge whether resources are appropriate and justified. Secondly, a cost that helped make the plan look complete is not automatically chargeable. Eligibility is tested during grant implementation, not awarded by a high score.
Horizon Europe eligible costs: the governing test
For most Horizon Europe actions, an eligible cost must be actually incurred by the beneficiary, incurred during the action period, necessary for the action, identifiable and verifiable in the accounts, compliant with applicable tax and social law, reasonable and justified, and consistent with the beneficiary’s usual accounting practices. The grant agreement and its annexes are decisive. The call text, work programme and particular action conditions may add restrictions.
This is why a budget should never be assembled by applying a percentage to the requested EU contribution and distributing the result among partners. Start with the work package, its tasks, the person-months required, purchases or access needed, and the delivery schedule. Then test each item against the grant rules.
The central question is not, “Could this organisation spend this money?” It is, “Can this beneficiary show that this cost was necessary for this action, allocated correctly and recorded in a way an auditor can verify?”
The main cost categories and where budgets fail
Personnel costs
Personnel is commonly the largest direct-cost category. It covers work done by employees or equivalent natural persons assigned to the action, calculated under the method set out in the grant agreement. The practical evidence trail normally runs from employment status and payroll through time records or a compliant declaration system to the task actually performed.
The recurrent weakness is a mismatch between narrative and budget. A work package may promise clinical validation across several sites but show too few person-months for protocol preparation, recruitment, data cleaning and analysis. Conversely, senior staff can be assigned implausibly high effort without a credible explanation of what they will do.
Evaluators are not conducting a financial audit, but they do see the implementation implication. If a task is critical, resource it explicitly. If a partner’s role is coordination, do not disguise delivery effort in a generic management line. The evaluator brief for the action type will determine how sharply this affects the quality and efficiency of implementation score, but unexplained effort is an avoidable opening for a significant weakness.
Subcontracting and purchase costs
Subcontracting is not a convenient category for work a beneficiary would prefer not to do. It is for action tasks outsourced to a third party where the beneficiary retains responsibility, the need is justified, and the arrangement complies with best-value and conflict-of-interest requirements. Core tasks should not be subcontracted without a strong and specific case. A proposal that gives a subcontractor an essential scientific or technical role while offering no rationale invites questions about consortium capacity.
Purchase costs cover items such as travel, equipment, other goods, works and services. These too must be necessary, reasonable and purchased in accordance with the beneficiary’s usual practice and the grant requirements. A named supplier in the proposal is not a procurement exemption.
The distinction matters in the narrative. External expertise contributing as a consortium beneficiary is different from a subcontractor delivering a defined task, which is different again from a supplier providing a service or item needed to perform the work. Blurring those roles damages both budget clarity and implementation credibility.
Equipment, depreciation and access
Buying a piece of equipment does not normally mean charging its full purchase price to the action. Where the equipment is used beyond the project or has a longer useful life, the eligible amount is generally the depreciation corresponding to the action period and the proportion of use for the action, subject to the applicable rules.
Applicants often underestimate the explanatory work this requires. A line for laboratory equipment, computing infrastructure or a pilot facility should be connected to a task, a period of use and an allocation logic. If the same resource serves commercial operations, teaching or another funded project, the action can claim only its defensible share.
In some cases, access to an existing facility, cloud capacity, testing services or specialised datasets is more proportionate than acquisition. That is not automatically preferable. The correct choice depends on the technical need, expected use and the beneficiary’s established practices. What matters to an evaluator is whether the choice is coherent with the methodology and timetable.
Financial support to third parties and other special cases
Some calls permit financial support to third parties, often called cascade funding, but only within stated conditions. The proposal must respect the call’s limits and explain the selection process, recipients, scope and controls. It is not a general contingency reserve.
Other categories, including internally invoiced goods and services, in-kind contributions, access to research infrastructure and specific additional-cost arrangements, require equally careful handling. Do not assume a category is available because it appeared in another Horizon project. Check the call documents and the version of the Model Grant Agreement proposed for the action.
Indirect costs are simple only after direct costs are right
Horizon Europe generally applies a 25% flat rate for indirect costs calculated on eligible direct costs, excluding defined categories such as subcontracting, financial support to third parties and certain excluded-cost bases. It is a simplification mechanism, not permission to stop thinking about overheads.
The usual error is mechanical calculation from the total direct-cost figure. That overstates the indirect-cost base where exclusions apply. The opposite error is to add a separate direct-cost line for routine administrative overhead already covered by the flat rate. Both create a budget that cannot be defended cleanly.
Keep the calculation transparent. Identify the eligible direct-cost base, remove excluded elements, then apply the rate. If the budget uses a special lump-sum or unit-cost model, follow the specific action rules rather than importing an actual-cost logic where it does not belong.
Eligibility is not the same as evaluator scoring
A cost can be eligible yet still weaken a proposal. Evaluators score award criteria, not future audit files. Depending on the action type, they assess Excellence, Impact and Quality and Efficiency of the Implementation against the published sub-criteria, thresholds and weighting. They will ask whether the resources match the work plan, whether the participants have the capacity claimed and whether the allocation of effort supports the pathway to results and impact.
That means a compliant-looking budget is insufficient. A €200,000 dissemination allocation will not cure a weak exploitation strategy. Nor will a large personnel figure persuade evaluators that a partner has the operational capacity to recruit participants, secure permissions or integrate a technical component. The budget must corroborate the proposal’s claims rather than compensate for their absence.
Read it in both directions before submission. From each work package, trace tasks to staff, purchases, subcontracting and deliverables. From each material budget line, trace back to a named task, beneficiary, output and timing. Any line that cannot survive both readings needs a justification, a revision or removal.
A pre-submission control that finds expensive contradictions
Before the coordinator freezes Part A and the technical annex, run a structured budget challenge. Test whether person-month totals reconcile across the work plan, tables and partner descriptions; whether subcontracting is justified and distinguished from procurement; whether equipment reflects depreciation and action use; whether indirect costs use the correct base; and whether every partner’s funding request is plausible against its role.
Then read the numbers as an evaluator would. Does the proposal allocate serious resources to the risks it says are critical? Does the coordinator have enough effort for governance, reporting and decision-making? Are small partners assigned delivery obligations out of proportion to their funded capacity? These are implementation questions, but they often expose weaknesses in methodology, risk management and consortium design as well.
A disciplined external review can help at this stage because the authors of a proposal tend to read intent into a number. An evaluator reads only the evidence on the page. BidShark’s automated assessment is designed to identify such evidence gaps against the applicable evaluation form; its score is not an audit opinion or a prediction of funding.
The useful final check is therefore not whether the total requested contribution fits the call envelope. It is whether every material euro tells the same story as the work plan, the consortium roles and the promised results. If it does not, resolve the contradiction before submission. After the deadline, the budget has to stand on its own.