SR&ED disputes rarely reach the Tax Court of Canada, but when they do, the decisions are unusually instructive, they show exactly where claims break down under scrutiny. The recent run of decisions, including Manning Canning, DAZZM, and Vortex Energy, repeats a pattern practitioners have watched for years.

The pattern: evidence beats testimony

In case after case, the court accepted that the taxpayer’s team did genuinely difficult work, and still denied the claim. Why? Because the taxpayer could not produce contemporaneous records showing a systematic investigation: hypotheses formulated, tests run, results recorded, conclusions drawn. Oral testimony describing work done years earlier, however credible, has consistently been found insufficient on its own.

The court’s framework comes from the five questions in Northwest Hydraulic (1998), which remain the operative test: Was there technological uncertainty? Were hypotheses formulated? Was the procedure consistent with the scientific method? Did the work seek a technological advancement? Were records kept as the work progressed?

That last question is doing the heavy lifting

Notice that record-keeping is not a bureaucratic add-on: it is one of the five questions. A company that solved a hard problem but documented nothing may hold an unprovable claim. The decisions make three practical points:

  • Records must be contemporaneous. Documents created during the work carry weight; reconstructions created for the review carry little.
  • Records must show iteration. A single design document proves intent. Versioned documents, test logs, and failure reports prove investigation.
  • Routine engineering is the boundary. Work that any competent professional could complete with standard practice, however commercially valuable, falls outside the program.

What this means for your next claim

The bar the court applies is the same bar a CRA reviewer applies, one step earlier. If your documentation would not survive the five questions, the time to fix that is during the year, not after a review letter arrives. Practically: keep dated technical notes, preserve failed branches and test results, and record why decisions were made, not just what was decided.

None of this requires heavyweight process. It requires habits, and they are exactly the habits we help clients build between claims.