CIR or CII Approval: Which One to Apply For, and Do You Need Both?

“We do innovation, so we will apply for CII approval.” The sentence sounds logical and is often wrong. The two approvals are not chosen according to how a company describes itself, but according to the exact nature of the work it performs for its clients and the tax scheme those clients claim. Here is the map, and the decision rule.

Two tax schemes, two logics

The starting point is not the approval but the tax credit your client will claim.

The research tax credit (crédit d’impôt recherche, CIR), set out in article 244 quater B of the French Tax Code, covers basic research, applied research and experimental development. Its central criterion is uncertainty: when the work starts, the state of the art does not tell you whether the result is achievable, or by what route. The rate is 30 % of eligible expenditure up to one hundred million euros, 5 % above, with a higher rate in the overseas departments. It is open to industrial, commercial or agricultural companies taxed on their actual profits, with no size condition.

The innovation tax credit (crédit d’impôt innovation, CII) is narrower on every count. It is reserved for SMEs within the meaning of EU rules and covers spending on the design of prototypes or pilot installations of new products. The key word is “new”: the product must stand apart from what the market offers through superior performance — technical, eco-design, ergonomic or functional. No scientific obstacle is required. Since 1 January 2025 its rate has been 20 % in mainland France, with an expenditure ceiling of 400,000 euros a year, that is a maximum credit of 80,000 euros; higher rates apply overseas and in Corsica. The scheme covers expenditure incurred up to 31 December 2027.

Remember the boundary: the CIR asks what we do not know how to do, the CII what the market does not yet offer.

What that changes for the provider’s approval

In both cases, a client can only claim a subcontracting invoice if the provider is approved. But the application does not go to the same place.

	Approval under the CIR	Approval under the CII alone
Authority	Ministry responsible for research	Ministry responsible for industry, via the Directorate General for Enterprise
What is examined	Capacity to conduct research work	Capacity to design prototypes or pilot installations of new products
Clients concerned	Any company eligible for the CIR	SMEs only
Filing window	1 January – 31 March for a first application; 15 August – 30 November for renewal	Reference date: 30 September
Duration	Three years as a rule; one year with no research operation in the previous twelve months; up to five years for long-standing holders	Five years

The linking rule is simple and saves a great deal of time: a company already holding a valid CIR approval does not need separate CII approval. The standalone CII application procedure is expressly aimed at companies that do not hold one. In other words, CIR approval is the most demanding and the most far-reaching: where it is attainable, it settles both questions.

The decision rule, in four questions

1. Does your work involve genuine scientific or technical uncertainty? Take three recent operations and write out, for each, the starting state of the art and the obstacle encountered. If the exercise holds up, aim for CIR approval.
2. Are your clients exclusively SMEs, and your deliverables prototypes of new products? If so, and if the previous exercise does not hold up, standalone CII approval is the right door.
3. Do you already hold valid CIR approval? Then file nothing more: your SME clients can claim their innovation tax credit spending.
4. What is today’s date? Trivial, and yet decisive. A CIR file lodged on 15 April will only take effect from the following year. So will a CII file lodged on 15 October. Our article on the approval timetable covers both campaigns.

One special case deserves a mention: style bureaux and stylists come under a separate procedure, that of the collection tax credit (crédit d’impôt collection), reserved for the textile, clothing and leather sector. The ministry lists them in a separate category of its public list of approved bodies.

What the academic literature says about these incentives

The question “do these schemes work?” is not rhetorical for a provider: it determines how stable its market is. Antoine Dechezleprêtre, Elias Einiö, Ralf Martin, Kieu-Trang Nguyen and John Van Reenen published a causal evaluation of R&D tax incentives in the American Economic Journal: Economic Policy in 2023, built on a change in eligibility thresholds and a regression discontinuity design applied to administrative data. They find a significant and persistent effect of tax relief on patenting and R&D spending, together with spillovers onto the innovation of technologically close firms (see the study).

The practical lesson is not to argue for the scheme, but to understand why states insist on traceability of the actors: the measured effect works precisely because the spending is real and carried out by capable teams. Approval is the administrative link in that chain.

Three framing mistakes that cost a year

Calling yourself “innovative” without looking at who your clients are. If your clients are mid-caps or large groups, the innovation tax credit does not concern them: they claim the CIR. Standalone CII approval is of no use to them.

Confusing difficulty with uncertainty. A long, costly and technically demanding development may involve no uncertainty in the tax sense, if the solution was known and only had to be applied. It is the most common ground for refusal on CIR files — and a legitimate reason to turn to the CII where the activity really is prototyping.

Waiting for a client to ask before filing. Between the filing window and the substantive assessment by a scientific expert, several months go by. A prospect asking for approval in September in order to sign in November cannot be satisfied that year if nothing was filed in the first quarter. Our article on subcontracted R&D and approval explains what that costs in lost business.

What if your activity covers both?

That is common among engineering firms: part experimental development for manufacturers, part prototyping for SMEs. The strategy is then clear: aim for CIR approval, which covers both needs. The file is more demanding — human resources, material resources, descriptions of research operations with state of the art and technical obstacle — but it is also the only one that opens the whole market to you. Our guide on what the ministry looks at in an approval file sets out the four blocks to prepare.

If the CIR file is not ready this year, standalone CII approval is an honourable step, valid for five years, that already secures your SME clients. Nothing prevents a later switch to the CIR, once a track record of research operations has been built.

Take action

Take twenty minutes and answer the four questions of the decision rule with your last three contracts in front of you. The choice almost always follows from that exercise — and it is better done in October, while both filing windows are still ahead, than in April once they have closed. Scheme status, procedures, frequently asked questions and a free ebook: it is all on our page dedicated to CIR-CII approval. To place this scheme among the other French recognitions, browse the full panorama of certifications and approvals.