Few areas of Canadian tax generate more friction between taxpayers and the Canada Revenue Agency than claims for Scientific Research and Experimental Development (SR&ED). The credit is generous, but the eligibility test is technical, and CRA reviewers frequently recharacterize genuine development work as “routine” engineering or ordinary “trial and error.” The Tax Court of Canada’s decision in Canafric Inc. v. The King, 2023 TCC 108 is a useful, accessible reminder of how the eligibility test actually works, and of how much can turn on credible witness testimony when the paper trail is imperfect.
What makes Canafric worth reading is its ordinariness. The taxpayer was not a semiconductor lab or a biotech company. It developed frozen pies. Yet Chief Justice Rossiter found that the work of reformulating those products to meet demanding client specifications involved real technological uncertainty and qualified as SR&ED. For any business that performs hands-on development without a formal research department, the decision is encouraging, and it sets out a roadmap for how to present an SR&ED file at the Tax Court of Canada.
The facts
Canafric Inc. is a food-manufacturing business that develops and produces custom frozen pies for major Canadian retailers. Its customers did not simply order existing products off a shelf. They came to Canafric with specifications, for example a pie that would hold a longer shelf life without artificial preservatives, that carried reduced salt or fat, that met particular nutritional or compliance targets, and that still delivered acceptable taste and texture after freezing and reheating.
To meet those targets, Canafric ran development projects, several each year, in which it formulated and tested new recipes and processes. The taxation years in dispute spanned roughly 2013 through 2016, and the appeal concerned a group of projects for which the company had claimed SR&ED expenditures and investment tax credits. The CRA reassessed to deny the claims, and Canafric appealed to the Tax Court.
The Agency’s central argument was familiar to anyone who has been through an SR&ED review. Its technical reviewer took the position that the work was essentially product development by trial and error, that any “breakthrough” achieved on one pie could simply be transferred to another, and that the projects therefore lacked genuine technological uncertainty. In the CRA’s view, this was cooking, not science.
The issue
The legal question was whether Canafric’s development work fell within the statutory definition of SR&ED in subsection 248(1) of the Income Tax Act, specifically within “experimental development” undertaken to achieve technological advancement. Because the statutory language is abstract, the Tax Court has long resolved that question through the framework set out by then-Associate Chief Judge Bowman in the foundational case Northwest Hydraulic Consultants Ltd. v. The Queen, 1998 CanLII 553 (TCC).
That framework is usually described as the “five questions,” and it remains the workhorse test in SR&ED appeals. Distilled, the Court asks:
- Was there a technological risk or uncertainty that could not be removed by routine engineering or standard procedures?
- Did the person formulate hypotheses specifically aimed at reducing or eliminating that uncertainty?
- Did the work proceed by the scientific method, including systematic observation, measurement, and the formulation, testing and modification of hypotheses?
- Did the process result in a technological advancement, that is, an advance in the understanding of the relevant technology (whether or not the project ultimately succeeded commercially)?
- Was a detailed record kept of the hypotheses tested and the results obtained as the work progressed?
The dispute in Canafric turned mainly on the first and fourth questions, whether there was true technological uncertainty and a resulting advancement, and on the fifth, the standard of documentation the taxpayer must meet to prove its case.
What the Court decided (and why)
Chief Justice Rossiter allowed the appeal, finding that the SR&ED claims at issue qualified.
On technological uncertainty, the Court rejected the CRA’s “just transfer the breakthrough” theory. The evidence, in particular the testimony of Canafric’s chief executive, established that a solution developed for one product could not simply be ported to another, because, as the Court accepted, the same ingredients react differently when combined in different products. Achieving a longer shelf life without preservatives in one pie did not tell the company how to achieve it in a different pie with a different filling, fat content, or moisture profile. There was no existing body of knowledge the company could consult to predict the outcome; it had to run experiments to find out. That is the essence of technological uncertainty under the first Northwest Hydraulic question.
On systematic investigation and the “trial and error” objection, the Court found that Canafric did not simply guess and re-guess. When a formulation failed to meet the client’s requirements, the company analyzed why it failed and adjusted the next iteration accordingly, an iterative, hypothesis-driven process rather than random tinkering. That distinction, between disciplined experimentation and mere trial and error, is frequently the decisive issue in SR&ED disputes, and the Court drew it in the taxpayer’s favour.
On technological advancement, the Court reaffirmed a point that taxpayers and reviewers alike sometimes forget: a project can advance technology even if it does not produce a marketable product or a commercial success. The advancement lies in the new knowledge generated about how to solve the technological problem, not in whether the resulting pie sold well.
On documentation, Canafric is especially valuable. CRA reviewers often treat the absence of perfect contemporaneous records as fatal. The Court confirmed that detailed, contemporaneous documentation is highly persuasive and should be kept, but that documentary evidence is not strictly mandatory. Credible, knowledgeable oral testimony, here from the company’s chief executive, supported by the records that did exist, can establish that the five-question test is met. The fifth Northwest Hydraulic question reflects sound practice and provides powerful evidence, but it is not an independent statutory bar that automatically defeats an otherwise eligible claim.
Why this decision matters / practical takeaways
Canafric does not change the law; it applies the long-settled Northwest Hydraulic framework. But the way it applies that framework offers several practical lessons for taxpayers facing an SR&ED denial:
- SR&ED is not limited to laboratories. Hands-on development in food, manufacturing, software, and the trades can qualify, provided the work confronts a genuine technological uncertainty that cannot be resolved by standard practice.
- “You could just transfer the result” is not a complete answer. Where a solution genuinely does not carry over from one product, material, or environment to another, the uncertainty is real, and the taxpayer should be ready to explain, in technical terms, why it does not carry over.
- Iteration with analysis beats iteration without it. The line between eligible experimentation and ineligible trial and error often comes down to whether each failed attempt was analyzed and used to refine a hypothesis. Contemporaneous notes that capture that reasoning are worth their weight in credits.
- Commercial failure does not disqualify a project. The test asks whether the work advanced technological understanding, not whether the product succeeded in the market.
- Documentation gaps are not necessarily fatal, but credible evidence is essential. A knowledgeable witness who can explain the hypotheses, the testing, and the results can carry the day, which is why preparing that witness, and assembling whatever records exist, is central to an appeal.
Because the taxpayer bears the burden of proof in a Tax Court appeal, the quality of the evidentiary presentation frequently decides these cases. Our guide to the role of evidence and the burden of proof at the Tax Court explains how that burden operates and why early preparation matters.
It is also worth keeping in mind that an SR&ED denial rarely travels alone. The same reviewer who recharacterizes development work as routine may also reduce related expenditures, adjust other parts of the return, or, in more aggressive files, propose penalties. Where the CRA goes further and alleges that a claim was made carelessly or knowingly without foundation, the analysis can shade into the territory of gross negligence penalties, which carry a separate burden on the Crown and their own evidentiary standard. Treating the SR&ED characterization, the quantum, and any penalty exposure as one coordinated problem, rather than three separate fights, usually produces a stronger and more coherent position on appeal.
How Barrett Tax Law approaches SR&ED files
SR&ED disputes are won or lost on the technical narrative. Our approach is to engage early, ideally during the CRA review or at the objection stage, so that the technological uncertainty, the hypotheses, and the experimental process are documented and articulated in the language the Northwest Hydraulic test demands, before positions harden.
When a claim has already been denied, we work to reconstruct the project history, identify the witnesses who can credibly explain the science, and map the evidence onto each of the five questions. Where appropriate, that work supports a strong objection; where the file proceeds further, it supports an appeal to the Tax Court of Canada. Many SR&ED denials arrive alongside broader CRA scrutiny, and our audit representation team can address the assessment as a whole rather than in isolation.
If your SR&ED claim has been reduced or denied, we offer a free, confidential consultation to review the file and discuss your options.
This article is commentary on a public decision of the Tax Court of Canada and is provided for general information only. It is not legal advice, and outcomes in SR&ED matters depend heavily on the specific facts and evidence of each case.
Frequently asked questions
What is the five-question SR&ED test from Northwest Hydraulic?
It is the framework the Tax Court of Canada uses to decide SR&ED eligibility, set out in Northwest Hydraulic Consultants Ltd. v. The Queen, 1998 CanLII 553 (TCC). The Court asks: (1) Was there technological uncertainty that routine engineering could not resolve? (2) Did the taxpayer formulate hypotheses to address it? (3) Did the work follow the scientific method? (4) Did it produce a technological advancement? and (5) Were detailed records kept of the hypotheses and results? Canafric Inc. v. The King, 2023 TCC 108, applies this same test.
What did Canafric Inc. v. The King decide?
In 2023 TCC 108, the Tax Court of Canada (Chief Justice Rossiter) allowed the taxpayer's appeal and found that its frozen-pie development projects qualified as SR&ED. The Court accepted that meeting demanding client specifications, such as longer shelf life without preservatives or reduced salt and fat while preserving taste, involved genuine technological uncertainty because ingredients react differently in different products.
Do I need perfect documentation to claim SR&ED?
Detailed, contemporaneous records are strongly recommended and are persuasive evidence under the fifth Northwest Hydraulic question. However, Canafric confirms that documentary evidence is not strictly mandatory. Credible, knowledgeable testimony about the hypotheses tested and the results obtained, supported by whatever records exist, can establish eligibility. The safest course is still to document the work as it happens.
Does 'trial and error' disqualify an SR&ED claim?
Not by itself. The key distinction is between disciplined, hypothesis-driven experimentation and random tinkering. In Canafric, the Court found the company analyzed why each failed formulation did not meet the requirements and adjusted the next attempt accordingly. That iterative, analytical process satisfied the scientific-method requirement and was not mere trial and error.
Can a project qualify for SR&ED if it failed commercially?
Yes. The Northwest Hydraulic test asks whether the work produced a technological advancement, meaning new knowledge about how to solve a technological problem, not whether the resulting product was a commercial success. A project that did not lead to a marketable product can still qualify if it advanced technological understanding.
What should I do if the CRA denies my SR&ED claim?
Act quickly, because Tax Court timelines are strict and the taxpayer bears the burden of proof. Reconstruct the project history, identify witnesses who can explain the technological uncertainty and the experimental process, and align the evidence with each of the five questions. Barrett Tax Law offers a free, confidential consultation to review a denied SR&ED claim and discuss objection or appeal options.