Case library

2022 FCA 72 (Rennie J.A.) affirming 2020 TCC 47 (Lafleur J.), docket 2017-3837(IT)G — appeal dismissed at both levels · Federal Court of Appeal (upholding TCC) · applied sciences / consulting

Why denied: Failed 4 of the 5 Northwest Hydraulic criteria; work did not follow the scientific method; expert evidence rejected for partisan tone + the taxpayer (not the expert) wrote most of the expert report.

Lesson: You cannot litigate around the Northwest Hydraulic test — it has been forcefully reaffirmed. Expert reports must be authored by the expert, balanced, and stay out of legal-conclusion territory.

Outcome: denied; FCA dismissed appeal

National R&D claimed SR&ED on "applied sciences" consulting work. CRA denied; TCC upheld the denial; National appealed to FCA arguing the TCC's reliance on the Northwest Hydraulic 5-question test was a legal error because the test "departs from the text of the ITA." **FCA forcefu

Source · retrieved 2026-08-05

2021 FCA 88 — docket A-442-19, affirming 2019 TCC 246 (Owen J.), docket 2016-4880(IT)G · Federal Court of Appeal · metal manufacturing

Why denied: Did not satisfy the Northwest Hydraulic 5-question test; specifically the scientific-method requirement.

Lesson: First major FCA-level rejection of the "Northwest Hydraulic departs from ITA text" argument. Foreclosed the test-attack strategy that National R&D tried again the following year.

Outcome: denied; appeal dismissed with costs fixed at $1,670

Kam-Press is the **predecessor case** to National R&D. The taxpayer made the same argument National R&D would make a year later — that the Northwest Hydraulic 5-question test exceeds the ITA's textual definition of SR&ED and should not be applied as a binding standard. **FCA reje

Source · retrieved 2026-08-05

Pattern across multiple TCC denials (Highweb & Page 2015 TCC 137, Hypercube 2015 TCC 65, Zeuter 2006 TCC 597, C.W. Agencies 2001 FCA 393) · Tax Court of Canada · cross-sector (most common in software, manufacturing)

Why denied: Work was the application of established techniques to a new context; difficulty + effort + skill ≠ technological uncertainty.

Lesson: A claim must name a specific technical question that competent professionals applying public knowledge could not answer.

Outcome: denied (most common single disallowance reason in TCC SR&ED jurisprudence)

This is not one case but the dominant pattern across denied TCC SR&ED claims. CRA reviewers and the TCC consistently disallow work that, while difficult and time-consuming, applied **known techniques** to a new context. The claim survives only if it can name a *specific technical

Source · retrieved 2026-05-09

Pattern across TCC denials; verified archetypes Highweb & Page 2015 TCC 137, Hypercube 2015 TCC 65, C.W. Agencies 2001 FCA 393 · Tax Court of Canada · cross-sector

Why denied: Hypotheses, experiments, and outcomes were not recorded contemporaneously. Reconstructed evidence at audit-time was insufficient.

Lesson: Q5 of the Northwest Hydraulic test ("was a detailed record kept as the work progressed?") is the dominant cause of audit failures even when Q1–Q4 are technically met.

Outcome: denied

If "routine engineering" is the dominant *narrative* failure, **Q5 (contemporaneous documentation) is the dominant *evidence* failure**. Many claims have genuinely novel work but die because the record wasn't kept *as the work progressed*.

Source · retrieved 2026-05-09

2001 FCA 393, [2002] 1 CTC 212 — FCA docket A-601-00 · Federal Court of Appeal (upholding TCC, Bonner J., docket 98-1324(IT)G, 30 Aug 2000) · software (in-house lottery order-processing system)

Why denied: In-house development of a large custom order-processing system (the International Distributed Lottery System) that the taxpayer genuinely could not buy off the shelf and did successfully build. The Crown's expert testified there was no uncertainty that could not be resolved by standard practice or routine engineering, so no technological advancement. Critically (para 10): the taxpayer retained no documentation, and called NOBODY personally involved in the development — only an expert who, lacking a project management plan, had to infer backwards from the finished product what problems must have been faced. Para 11: the unexplained failure to call the project manager supported an adverse inference.

Lesson: This is the case that canonised the Northwest Hydraulic 5 questions at the appellate level (para 17) — every later decision quotes CW Agencies' formulation. Its facts are the direct warning for a software portfolio: 'we needed it, nothing off the shelf did it, we built it, it works' is a BUSINESS narrative, not a technological-uncertainty narrative. And the people who did the work must testify; a reconstructing expert is not a substitute.

Outcome: denied; TCC dismissed with costs, FCA dismissed the appeal

CW Agencies marketed government lottery tickets worldwide — 2-3 million customers, up to 25,000 purchase orders a day. No commercial software met its requirements for flexibility, scalability, instant response, accuracy and reliability, so it built the IDLS in-house and succeeded

Source · retrieved 2026-08-05

2006 TCC 597, 2007 DTC 41 — docket 2005-3306(IT)I (Informal Procedure) · Tax Court of Canada · software (web-based educational tool for high-school students)

Why denied: The Alien Travel Guide was found useful, valuable, and genuinely non-existent before the taxpayer built it — and still ineligible. The taxpayer used commercially available application programming tools to create a computer-aided instruction program; no technological uncertainty was shown that standard engineering could not overcome. Supporting documentation was inadequate (para 28).

Lesson: The single most important software holding in the library: NOVELTY IS NOT ADVANCEMENT. A product that did not exist before, that is commercially valuable and technically competent, is still not SR&ED if it was assembled from generally available tools. Also settles two recurring taxpayer arguments: (1) CRA allowing the claim in prior years binds nothing — each tax year stands on its own (para 13, citing Lunn); (2) reliance on another government body's statement of eligibility (here FedNor) is not a defence.

Outcome: denied; appeals dismissed without costs

Zeuter's sole shareholder, an engineering-physics and electrical-engineering graduate, built the 'Alien Travel Guide' — an online learning tool for high-school physics and mathematics. Over ten years and 70+ summer students it produced 20,000+ web pages covering 1.5 million subje

Source · retrieved 2026-08-05

2015 TCC 65 — docket 2014-1254(IT)I (Informal Procedure) · Tax Court of Canada · software (web-site source-code analysis / crawler tooling)

Why denied: A five-employee IT shop built a crawler-based tool to read and analyse website source code and produce a PDF diagnostic. The court found every difficulty was resolved with recognised programming techniques, using commercially available tools and existing languages — the founder himself conceded on cross that crawlers are a commonly used tool (para 45). The problems described (domain-name handling, browser compatibility, a coding error above 150 runs/day, URL redirects) were normal programming problems (paras 23-26, 46). Documentation was a program tree diagram never explained in court and an hours log that 'does not appear to reflect reality' (para 48).

Lesson: THE most directly applicable failure case for a solo/small-team software claim. Three lethal details, all reproducible in this portfolio: (1) the claim covered wages of three developers but the time log recorded only the founder's hours (paras 10, 27) — the two employees with the actual computer-science backgrounds never testified; (2) the listed obstacles were debugging, not uncertainty; (3) 'trial and error' was described as the method, which is not the scientific method. Note the court let a CRA advisor testify as a fact witness about eligibility even after doubting his expert qualification (para 30).

Outcome: denied; appeal dismissed ($28,800 of SR&ED expenditures disallowed)

Hypercube, a five-person Montreal web-programming firm, claimed $28,800 for developing a program that crawls a website's source code, analyses it for weaknesses and performance, and emits a PDF diagnostic report. Founder Karl Villeneuve (BBA + MBA, some CS coursework) was the onl

Source · retrieved 2026-08-05

2015 TCC 137 — docket 2014-1703(IT)I (Informal Procedure) · Tax Court of Canada · software (iFactum cross-platform web application, .NET / Java / J# / JavaScript)

Why denied: Porting and integration work across .NET/Java/J#/JavaScript was found to be selective trial and procedural sequencing — 'knowing which available products applied in the correct sequence utilizing routine, standard or customary modifications' (para 18) — not a knowledge gap. The two hypotheses were only assembled AT THE HEARING, under questioning from the Bench, and it was not clear they had been documented when the work was done (para 20). The witness had to describe the work almost entirely by reference to CRA's own Technical Review Report, which was written AFTER the work (para 22).

Lesson: The most useful case in the library for proving what contemporaneous documentation is worth, because it contains its own control. On identical technology, in the same claim, ONE sub-project was allowed — and the court explains why: 'Considerably more recordal information was generated' for it (para 19). Same company, same year, same tech stack; the only variable that moved was the record. The court's language is unusually direct: 'detailed step-by-step logging, analysis, and measurement is a mandatory requirement, not an optional addendum. It is the roadmap.' Note also that the allowed sub-project produced a NEGATIVE result and still qualified — knowledge gained by disproving the hypothesis.

Outcome: PARTIAL — the Minister had already allowed one sub-project (STA2 Phase II, $2,704 of ITCs); the appeal for everything beyond it was dismissed ($25,200 disallowed in 2007, $37,975 in 2008)

Most documentation cases show a claim failing for want of records. Highweb shows a claim *partly succeeding* for want of nothing else — which makes it the rare case with an internal control.

Source · retrieved 2026-08-05

2025 TCC 63 — dockets 2019-894(IT)G, 2021-1471(IT)G · Tax Court of Canada · energy services (mobile direct-contact water heaters for hydraulic fracturing)

Why denied: The taxpayer built truck/trailer-mountable high-efficiency direct-contact water heaters for the fracking industry — the only units of their kind in Alberta's oil patch — and one unit failed catastrophically during testing. Claimed $1,454,874 for 2014. Still denied: the work was not shown to be experimental development within s. 248(1). Spiro J. found the tests were conducted by trial and error, i.e. no systematic investigation, adopting Rossiter C.J.'s statement in Canafric (2023 TCC 108 at para 101) that trial and error used alone does not come within the scientific method because it aims to solve a functional problem using different options.

Lesson: The most recent full-length TCC treatment of the five criteria and a caution against over-reading Canafric. Vortex's counsel argued Canafric heavily; the court still dismissed. Also a warning about outcome-based intuition: a genuinely first-of-its-kind machine, a catastrophic test failure, and a seven-figure spend are not themselves evidence of technological uncertainty pursued by scientific method.

Outcome: denied; appeals for 2014 and 2019 taxation years dismissed with one set of costs

Vortex tried to build high-efficiency, direct-contact water heaters mountable on a truck or trailer for the fracking industry. It claimed $1,454,874 of SR&ED for 2014 (work on two ten-million-BTU units, the first of which failed catastrophically during testing) and further amount

Source · retrieved 2026-08-05

2025 TCC 129 — docket 2019-3604(IT)I (Informal Procedure) · Tax Court of Canada · construction (mobile roller-compacted concrete)

Why denied: Paveit, a 10-employee Alberta paving company, field-tested another company's prototype concrete mixer and claimed its contribution as process development. It failed criteria 2, 3 and 5: no technological uncertainty was properly ascertained, so no specific hypothesis could be aimed at one (paras 46, 49 — 'lack of rigor'); the procedure did not accord with the scientific method; and the documentary evidence 'was lacking' (para 57) — tables of test results without the reasoning connecting them.

Lesson: Newest decision in the library and the cleanest statement of a cascade failure: if the uncertainty is never properly identified, criteria 2, 3 and 4 cannot be satisfied either — 'If a technological uncertainty was not ascertained, it is difficult to see how specific hypotheses could be formulated' (para 46). It also confirms the taxpayer-favourable rule that an advancement may consist of REJECTING a hypothesis (paras 52, 55, following Joel Theatrical Rigging) — but only where the record shows the hypothesis and the rejection. Relevant to collaboration claims: doing the field testing for someone else's prototype is not automatically your own SR&ED.

Outcome: denied; appeal from the 2017 assessment dismissed without costs

Paveit is a 10-employee paving company in Slave Lake, Alberta. Its 2017 claim related to a Mobile Roller-Compacted Concrete project. An unrelated company, Rock Solid, had received SR&ED credit for developing the mixer used; Paveit argued that its field testing of that prototype,

Source · retrieved 2026-08-05

97 DTC 99 (TCC) — docket 91-2556(IT)G, 28 Aug 1996 · Tax Court of Canada (SET ASIDE by the FCA in 1998) · construction (concrete forming / spatial membrane panel)

Why denied: The Tax Court accepted that testing had been undertaken and that a technological advance in the construction industry had resulted, but held the claim failed a 'repeatability' criterion: research is not systematic unless records permit others to repeat the exact tests. The taxpayer's documents were at the deceased engineer's residence and never produced. The FCA rejected this reasoning in 1998 and allowed the taxpayer's appeal.

Lesson: Included for one reason: to stop the citation error. The RIS-Christie name attaches to two decisions with opposite outcomes, and the strict documentary-repeatability rule people attribute to 'RIS-Christie' is the TCC holding that was OVERTURNED. Citing it as good law against a taxpayer — or filing the FCA decision as a taxpayer loss — misstates the authority. Practical takeaway from the reversal: records still matter enormously, but their absence is not automatically fatal where testing and advancement are otherwise established.

Outcome: SUPERSEDED — 1982 year allowed by partial consent; 1983 year dismissed. The SR&ED holding was set aside by the Federal Court of Appeal in 1998; see 04-ris-christie-1998-fca.

Not as precedent — as a disambiguation guard. 'RIS-Christie' is cited constantly in SR&ED practice, and the proposition usually attached to it (that systematic investigation requires documentation permitting exact repetition) comes from THIS decision, which the Federal Court of A

Source · retrieved 2026-08-05

2021 TCC 17 — docket 2018-528(IT)G (General Procedure) · Tax Court of Canada · construction (energy-efficient / passive house)

Why denied: The appellant (carrying on business as Global Sustainable Solutions) built a bungalow in southern Ontario intended to run without grid-tied energy. Its counsel stated the hypothesis in oral submission as 'I can build a house ... that doesn't need a furnace, at a price point that's similar to a regular home' and identified the technological uncertainty as 'the cost' (para 26). The Court disagreed that cost is a technological matter at all: 'Conceptually there is no technological aspect implicit in the notion of an item costing or priced at x rather than y dollars' (para 30). None of the construction techniques described — extra footing/wall/attic insulation, taped seams, ICF walls, a cheaper ground loop — was said to be new technological knowledge (para 27), and no expert evidence was led as to the state of knowledge in the industry (para 28). The only testing (three recycled-content concrete mixes) was 'entirely ... routine engineering, and not based on any identified hypotheses. Trial and error' (para 34). No detailed record of hypotheses and results was kept (para 36).

Lesson: A COMMERCIAL objective is not a technological uncertainty, and restating a cost target as a hypothesis does not convert it into one. Para 31 is the sentence to remember: an hypothesis 'specifically aimed' at reduction of a cost or price 'quite misses that Northwest Hydraulic target'. Para 33 supplies the narrow path that was not taken here — uniquely combining several known procedures CAN be SR&ED, but only where the combination is itself achieved through the scientific method and produces new technological knowledge, proved by evidence. Note also para 35: a failed attempt still counts as advancement if it yields new technological knowledge — the defect was the absence of knowledge, not the absence of success.

Outcome: denied; appeal of the 2012 and 2013 loss determinations dismissed, with costs

The corporate appellant, carrying on business as Global Sustainable Solutions, appealed two loss determinations for its 2012 and 2013 taxation years denying SR&ED expenditures (reduced at the hearing to $139,631.67 and $61,752.60) incurred building a small, highly energy-efficien

Source · retrieved 2026-08-05

2017 TCC 179 — docket 2015-2330(IT)I (Informal Procedure) · Tax Court of Canada · food and beverage manufacturing (plant sterol beverages, bottle filling)

Why denied: Two projects, both denied. Project 705 (Plant Sterols Beverage) attempted to disperse plant sterols in a beverage using methods already known in the industry — no technological uncertainty that routine engineering could not remove, no hypothesis formulated, and no technological advancement. Project 806 failed first on identification: it 'was presented in a manner that was so confusing that it was impossible for me to determine if that project had to do with the dual-chambered bottle or the Partial Hot Fill System during the years under appeal' (para 75). Taken at its best as the Partial Hot Fill System, it still failed — 'not materially different from the hot fill process widely used in the beverage industry', 'designed using standard practices and routine engineering' (para 72), with no evidence led on what gap existed in the market (para 70) and again no hypothesis (para 74).

Lesson: The purest 'what did you actually claim?' failure in the library. Before any criterion is reached, the claimant must be able to say which project is under appeal and what it consisted of — Flavor Net could not, and para 75 turns that alone into a dispositive finding. Two further reusable points: (1) para 70 — you must lead evidence of the STATE OF THE MARKET/industry to show the gap your work addressed; asserting novelty without that baseline proves nothing; (2) the appellant's own witness described the result as 'an alternate to the existing hot fill standard that worked for our application' (para 71), which the Court read as adaptation of known technology to a private use case, not advancement. A thing that is new TO YOU is not new to the field.

Outcome: denied; appeal dismissed for both projects, without costs

Flavor Net appealed reassessments denying SR&ED expenditures and related ITCs for its taxation years ending June 30, 2010 and October 31, 2010. Two projects were at issue: project 705, the Plant Sterols Beverage Project, and project 806, which the appellant was unable to consiste

Source · retrieved 2026-08-05