Research tax credit (CIR): 2026 guide

Updated March 2026 - The French research tax credit (CIR) remains one of the most significant innovation support mechanisms available to businesses in France. But it is also one of the most closely scrutinised — and most frequently misunderstood. The real challenge is not "doing CIR" as a routine exercise, but determining whether your work genuinely qualifies as éligible R&D, and then building a technical and financial file strong enough to withstand scrutiny.

The CIR is a tax credit granted to companies that incur certain qualifying research expenditure. It covers fundamental research, applied research and experimental development, provided the statutory criteria are met. It can be offset against corporate income tax or refunded directly to éligible companies (particularly SMEs and innovative businesses).

Under the applicable legislation and tax doctrine:

• researcher and research technician salaries and their social charges, weighted by the proportion of time dedicated to éligible R&D;
• operating costs calculated as a fixed percentage of éligible personnel costs;
• depreciation of assets specifically assigned to éligible R&D activities;
• subcontracted R&D assigned to accredited public bodies, universities or certain approved private organisations, within the conditions and caps provided for.

The greatest risk in a CIR file is not an arithmetic error — it is a wrong qualification of the project. To establish eligibility, you must demonstrate:

• the existence of a genuine scientific or technical uncertainty at the outset of the work;
• the systematic methodology used to address or resolve it;
• the concrete work carried out — experiments, tests, iterations, documentation of failures;
• the means deployed — human resources, equipment, time allocation.

The CIR connects directly to our article on public innovation funding 2026, the guide on financing your SME in 2026 and our CIR, CII and JEI advisory service.

A serious CIR file must cover all of the following:

• a description of each project in technical, not commercial, terms;
• demonstration of the innovative or research character — not just that the outcome is new, but that it was uncertain at the start;
• a timeline of activities with milestones and results;
• human resource allocation logs — who worked on what, for how many hours, and with what justification;
• the detailed financial calculation, reconciled with payroll records and invoices.

Hayot Expertise advice: a solid CIR file is built during the project, not retrospectively when preparing the tax return. Documentation constructed after the fact is always weaker and always more exposed in an audit.

• confusing commercial innovation (new product for the market) with R&D eligibility (genuine scientific or technical uncertainty) — these are not the same;
• overstating the time devoted to éligible research activities in personnel allocation;
• documenting only the financial side of the claim without corresponding technical evidence;
• ignoring the technical justification file entirely, or producing it only in summary form.

We can help you qualify the work, structure the supporting evidence and secure the filing.

Discover our CIR, CII and JEI advisory support

A good CIR file does not start at filing time. It is built as the project moves forward, while hypotheses are tested, obstacles appear and the team records what was actually tried. That traceability is what later proves that the work falls within research or experimental development, rather than ordinary improvement work.

Separate research, innovation and adaptation#

The most common mistake is to assume that an innovative project is automatically éligible. In practice, CIR requires more than a good idea. You need to show a scientific or technical uncertainty, a working method, hypotheses, tests and real progress toward a solution. A commercial or marketing innovation may be valuable to the business, but it does not always open the door to the tax credit.

What the technical file should tell#

The file should let the reader follow the project from the starting point to the result achieved. It should explain:

• the context and initial state of the topic
• the technical or scientific difficulty encountered
• the approaches explored to solve it
• the tests carried out and the conclusions reached
• the part of the work actually devoted to research

This narrative should be easy to read. A strong file does not pile up forms; it explains the project's logic with concrete evidence. That coherence is what makes the treatment safer during an audit or an internal review.

The financial side has to follow the same logic#

The financial calculation must stay aligned with the technical proof. Time spent, payroll costs, depreciation, subcontracting and related expenses must be traceable to a clearly identified project. If the link is weak, the risk of challenge increases.

CIR also fits into a broader ecosystem: it should be connected to public innovation support in 2026, to the overall financing strategy of the SME and to CIR, CII and JEI support when the project scales up.

Anticipate the audit instead of reacting to it#

The safest way to secure CIR is to anticipate the questions an auditor would ask. Who worked on what? When? For what objective? What was attempted? Why was the final solution not obvious? If the file already answers those questions, it is much stronger.

Common mistakes that weaken the file#

• confusing product improvement with éligible research
• documenting the finance side only and leaving out the science or the technical path
• forgetting to date tests and iterations
• overstating time spent without a reliable tracking basis
• waiting until closing to rebuild the evidence

<details> <summary>Is CIR only for large companies?</summary>

No. The scheme can also apply to SMEs and startups, provided the work and expenses meet the criteria. Company size is not the deciding factor.

</details> <details> <summary>Do you still need a technical file for a small claim?</summary>

Yes. The level of proof should be proportionate, but the principle is the same: explain the project, the work carried out and the logic of eligibility.

</details> <details> <summary>Can the file be rebuilt after the fact?</summary>

Partly, but it is not ideal. The more the evidence is collected in real time, the more credible and easier to defend the file becomes.

</details>

In 2026, the CIR remains a powerful lever for funding R&D, provided it is treated as a complete technical and financial file rather than a routine claim. The earlier the documentation is built, the more defensible the position becomes.

(Official sources: Service-Public on the research tax credit, Ministry of Higher Education CIR guide, BOFiP on the research tax credit)