IRAP Funding for Canadian Corporations: Get Up to 80% of Your R&D Salaries Back

If your company is doing real research and development and paying for all of it yourself, you may be leaving a cheque on the table. Canada's Industrial Research Assistance Program (IRAP) covers a large share of that cost, and unlike a loan, you never pay it back.

The National Research Council (NRC) runs the program. It gives Canadian small and mid-size businesses non-repayable contributions toward projects that push past a genuine technical problem. The money arrives as cash while the project runs, not as a tax credit you claim next spring. For a growing company with a hard engineering problem and a tight balance sheet, that timing changes what you can afford to attempt.

This post covers the small and mid-size project streams, what makes a project fundable, and how the numbers land.

The small and mid-size project streams

IRAP funds projects at two sizes that matter to most companies.

The smaller of the two is the Accelerated Review Process (ARP). It contributes up to $50,000 toward a focused project. You finish the work inside one fiscal year, which runs April to March. Approval moves faster than the larger stream, so it suits a well-defined piece of work you want to start soon.

Above it sit the mid-size projects. These run from roughly $100,000 to $500,000 or more. A mid-size project can stretch over as long as 36 months, so it fits sustained development work: a new production process, a novel material, a control system you have to invent rather than buy.

Both streams share the same test for what counts, and the same way of paying you.

What makes a project fundable

The program does not fund products, and it does not fund work just because it is expensive. It funds the resolution of a technical uncertainty. Your project needs three things on paper.

First, a hypothesis. You are trying to do a specific thing and you have a theory for how it might work. Second, genuine uncertainty, meaning the outcome is not something a competent engineer in your field can already predict or look up. Third, a clear baseline: a statement of what is possible today, and why the thing you want cannot be done with what already exists.

Advanced manufacturing is a natural fit. Say you run a shop that fabricates precision components, and you want to hold a tolerance your current process cannot reach without scrapping half the batch. Nobody sells a machine that solves it, and the trade literature has no answer. So you form a hypothesis about tooling and feed rates, and you start running experiments.

That is an IRAP project. The baseline is your current scrap rate. The uncertainty is whether the new approach can hold the tolerance at volume. The hypothesis is the change you believe will get you there.

The pattern holds across software, materials, chemistry, and hardware. If a knowledgeable person in your field would have to experiment to find the answer, you are likely in scope.

The money

Here is how the contribution works once a project is approved.

The program covers up to 80% of the salaries of your technical staff on the project. It covers up to 50% of the cost of Canadian subcontractors you bring in for specialized work. Materials consumed in the work can qualify too. On larger mid-size projects the labour share sometimes lands in the 65% to 80% range rather than at the top, because you negotiate the contribution with an Industrial Technology Advisor.

The payment model is the part companies tend to like. You do the work, submit a claim for the period, and the program reimburses its share. NRC pays fast. In its 2024 to 2025 year it met the 35-day standard on almost every claim, and turned the average one around in a few days. So the funding arrives in step with your spending rather than a year later.

There is one ceiling to plan around. Total government support for a project caps at about 75% of eligible costs, counting every federal, provincial, and municipal source together. Your company funds the remaining quarter from its own resources. IRAP will never cover the whole bill, by design.

IRAP and SR&ED together

The 75% cap matters most when you combine IRAP with the other big federal program, SR&ED. The two work in sequence rather than compete.

IRAP pays you during the project, against your claims. SR&ED credits arrive after your fiscal year, when you file. An IRAP contribution counts as government assistance, so it reduces the expenditure base under your SR&ED claim. You do not get to count the same dollar twice. Even so, the combination is usually the strongest funding position a Canadian small business can build: cash in during the work from IRAP, a refundable credit after it from SR&ED, both under the same 75% ceiling.

The catch is the same one SR&ED has. The story of the work has to exist as a record: the hypothesis, the baseline, the experiments, and the results. That record is far easier to keep as you go than to reconstruct a year later.

Where XY Space fits

We build custom AI systems, often for companies in manufacturing and other operational settings, and that work regularly clears the IRAP bar on its own terms. The hard parts are real technical uncertainties rather than routine integration.

So we scope our engagements around the funding as well as the build. We help you frame the hypothesis and baseline the way the NRC reads them, split the work into claimable milestones, and keep the technical record that both IRAP and SR&ED depend on. The result is a project that ships and a funding position that holds up.

If your company is spending its own money to get past a hard technical problem, some of that spend is very likely fundable right now. Talk to XY Space about what you are planning, and we will help you build it so the innovation and the funding come out of the same project.