Software development qualifies for SR&ED in Canada when it clears three specific tests the CRA applies to every claim. Not "we built something impressive." Not "we used cutting-edge tech." Three tests, and your project has to pass all of them. Here they are in plain English, with the Canadian rules that decide what you get back.
Software development qualifies when the work sought a technological advancement, faced technological uncertainty that standard practice couldn’t resolve, and was carried out as a systematic investigation. The CRA looks for all three together. Miss one and the project, or that part of it, falls out of the claim.
The CRA isn’t paying you for the software. It’s paying you for the knowledge you had to generate because the existing knowledge ran out. If you can point to the moment the manual, the framework docs and the senior engineer all said "nobody knows, it has to be tried," you’re in SR&ED territory.
At the start, you couldn’t know whether your goal was achievable, or how to achieve it, using existing knowledge and standard practice. This is the test software teams most often fail to show, because the uncertainty feels obvious in the moment and then gets written up as a finished feature. Business uncertainty (will people buy it) does not count. Technical uncertainty (can it be made to work) does. A quick self-check: could you have looked up the answer, hired someone who already knew it, or bought a product that did it? If yes, there was no uncertainty in the SR&ED sense.
You worked through the uncertainty in a structured way: you formed a hypothesis, tried it, measured the result, and adjusted. It doesn’t have to look like a science experiment. For a dev team it usually looks like a branch, a prototype, a benchmark, a failed approach, and a second approach. The trail of that work is what proves it happened.
You were trying to move past the standard practice available in your field, even by a modest amount. The advancement has to extend the technology beyond what a competent professional could readily deduce. It doesn’t have to be a world first, but it can’t be knowledge you could have looked up or bought. And you don’t have to succeed: failed and abandoned work qualifies on the same terms, because you still generated knowledge about what doesn’t work.
Software work does not qualify when the outcome was never in real technical doubt, even if it took months and a strong team to finish. Hard is not the same as uncertain. Work that usually does not count on its own:
One caveat: a test harness built to run your experiments, or data-handling code that exists only to trial the uncertain part, can qualify as support work tied to the uncertain core. On its own it does not.
There is no CRA list of approved software; no category is automatically in or out. What qualifies is the work, in any domain, that resolves a genuine technical unknown, and it most often appears here:
Systematic investigation for a software team is the ordinary loop of hypothesis, experiment, and measurement, captured well enough that someone else can see it happened.
Your commits, pull requests, benchmark logs, design docs and stand-up notes already tell this story. If you did not track hours, you can reconstruct time from them and from tickets and release history, though a contemporaneous record is stronger and easier to defend. Fix it going forward by tracking the technical work as you go.
The enhanced refundable credit goes primarily to Canadian-controlled private corporations, or CCPCs. If that’s you, the first $6 million of qualifying expenditures earns a 35% credit that’s refundable, meaning the CRA pays it to you in cash even if you owe no tax. That’s up to $2.1 million a year, though the enhanced rate phases out as taxable capital grows and the limit is shared among associated corporations. The $6 million limit applies to taxation years beginning on or after December 16, 2024, the change Bill C-15 made.
Non-CCPC corporations generally earn a 15% non-refundable credit, which reduces tax owing rather than paying out as cash. From the same taxation years, eligible Canadian public corporations earn the enhanced rate as well.
A CCPC building a fraud-detection engine, chasing an accuracy and speed target nobody could promise up front, has a qualifying base of about $690,000. At the 35% enhanced federal rate that earns roughly $241,000 in federal SR&ED credit before the prescribed-proxy overhead amount. The eligible work broke down like this:
Two more categories count when they apply: materials consumed or transformed in the work, and overhead, claimed as actual costs or through the prescribed-proxy amount, which is 55% of eligible salaries. Run both overhead methods and take the higher.
Because it’s a CCPC, that federal credit is refundable: it comes back as cash even in a loss year, which matters most to a pre-revenue software company. Most provinces and territories run their own R&D credit as well, and where one exists it makes your total return larger. How much larger depends on the province. The CRA lists them in its provincial and territorial R&D credits table, and the State of SR&ED overview gives the bigger picture.
A project qualifies on its technical uncertainty, so frame it around that, not around the product.
The gap isn’t between companies that innovate and those that don’t. It’s between companies that recognize their innovation and those that don’t. Framing is how you recognize it.
Software development qualifies for SR&ED in Canada when you can show the uncertainty, the investigation, and the advancement, and when you file inside the window as the right kind of corporation. Get those right and the credit follows. If you want help pressure-testing whether your project qualifies, grab a free consultation and we’ll go through it with you. The CRA’s SR&ED program page has the official criteria if you want to read the source.
Related reading: how SR&ED tax services support startup R&D credits, first-time SR&ED claim help in Canada, and the common SR&ED claim mistakes that cost refunds.
No. For a Canadian-controlled private corporation, the enhanced 35% credit is refundable, so it’s paid to you in cash even in a year you make no profit and owe no tax. That’s a big reason SR&ED matters so much to early-stage software companies.
No. A normal development team qualifies when its work meets the three tests. What matters is technological uncertainty and systematic investigation, not job titles or a dedicated research department. Team size does not change eligibility either; it only changes the dollar amount, through the salaries claimed.
No. Internal-use software can qualify if it meets the same three tests. The claim rests on the technical uncertainty you resolved, not on whether the software became a product or generated revenue.
Often yes. Arm’s-length Canadian contractors are claimable at 80% of the eligible amount when the work is done in Canada on your behalf by a taxable supplier. Their work has to meet the same eligibility test as in-house work.
No. Almost everyone builds on existing tools. The question is whether combining or extending them forced you past a technical unknown that had no obvious answer. Standard configuration does not qualify. Pushing a platform past its documented limits, or making incompatible systems work together with no recipe to follow, can.
The CRA’s service standards are 60 days for refundable claims accepted as filed and 180 days for refundable claims selected for review, measured from a complete claim. Those are targets, not guarantees, and a clean, well-documented claim is the best way to stay on the faster track.
This article is general information, not tax advice. Tax figures depend on your corporation type, province, and taxation year.
Do you have a SRED question? Planning for the future or perhaps you want to know how much your claim might be? Don’t worry, our CPA is always ready to answer any question. Get a SRED expert in your corner.
Have a question? We’d love to help. If you don’t have a SR&ED expert in your corner, doesn’t it make sense to have one?
