The SR&ED definition and the CRA's two questions, Why and How, apply identically in every sector. What differs is where the technological uncertainty tends to sit, which exclusions bite, and what the CRA's reviewers have learned to look for. This page walks through the main sectors in Canadian claims; the industry guides go deeper on each.
Where qualifying work sits by sector
| Sector | Usually SR&ED | Usually not | Watch for |
|---|---|---|---|
| Software | Experimental development to resolve technological uncertainty in algorithms, architecture, performance or integration where standard practice did not provide the answer | Feature development with known techniques, configuration, routine debugging, UI design | Technological versus business uncertainty; agile records |
| Manufacturing | New or improved materials, processes and products developed through experimentation; experimental production | Commercial production, quality control, routine tooling | Experimental production versus commercial production with experimental work |
| Life sciences | Basic and applied research, preclinical and clinical experimental development, formulation and device development | Routine testing, regulatory paperwork, marketing studies | Clinical trial contract payments; work outside Canada |
| Agriculture and food | Breeding and growing trials, process and formulation experiments with measured outcomes | Normal farming, style changes to recipes, routine data collection | Materials transformed into saleable product; recapture |
| Natural resources | New extraction, processing, monitoring or remediation technology tested experimentally | Prospecting, exploring, drilling for or producing minerals, petroleum or natural gas | The resource exclusion |
| Cleantech and energy | Materials, storage, generation and grid experiments | Installing proven systems | Government assistance reducing qualified expenditures |
Software
Software is the largest and most reviewed sector. The CRA's position is that software development is SR&ED when there was a technological uncertainty that could not be resolved by applying standard practice, and the work proceeded as a systematic investigation. The common failure is confusing a business or product uncertainty, such as whether a feature would be adopted or how to meet a specification, with a technological one. Novel combinations of known components, integration of systems through documented interfaces and performance tuning by standard methods are development, not SR&ED. Where a team did not know whether an approach could achieve a performance, scalability or accuracy target, and tested alternatives systematically to find out, the work qualifies, and the extent of eligible work is the experimentation, not the whole release.
Agile teams meet the How question by tagging tickets and commits to the uncertainty they address and recording what each sprint learned. The RTA will ask when the uncertainty was resolved; SR&ED ends there, and the remaining release work is commercial.
Manufacturing and engineering
Manufacturers hold eligible work on the shop floor that they rarely call research: a new alloy through an existing line, a process parameter study to reach a tolerance, a fixture that had to be developed and tested. Each is experimental development if the outcome was uncertain and tested systematically. The CRA's policy on developing an asset draws the line that matters most: experimental production, where the purpose of the run is to evaluate the experimental development, is SR&ED, and its full cost can be claimed; commercial production with experimental development work is not, and only the incremental cost of the experimental work qualifies. Materials transformed into product that is later sold trigger recapture of the credit.
Life sciences
Pharmaceutical, biotechnology and medical device development is the paradigm of SR&ED, and the issues are structural. Clinical trials run by contract research organisations are contract payments: 80% of arm's length payments count, and only for work performed in Canada, so offshore trial sites earn no credit. Work with universities and hospitals is often a third-party payment, requiring the claimant to be entitled to exploit the results. Many life sciences companies are CCPCs with no revenue, so the refundable enhanced credit is central to their funding, and the taxable capital test matters as they raise money.
Other sectors
Agriculture and food processing. Field and growing trials, feed studies and process experiments qualify where the outcome was uncertain and measured. The CRA's agriculture guidance warns against claiming normal production; crops or animals from trials that are sold are materials transformed, so recapture may apply. Recipe changes for taste are style changes; formulation to reach a technical result such as shelf life can qualify.
Natural resources. Prospecting, exploring, drilling for and producing minerals, petroleum and natural gas are excluded. Developing a new extraction, processing or monitoring technology and testing it experimentally is not, and the resulting technology's trial use can be experimental production.
Cleantech and energy. Development qualifies in the ordinary way. Government assistance for the same work, which is common in the sector, reduces qualified expenditures and must be tracked by project.
Construction and engineering services. Novel structural, materials or systems solutions can qualify, but the company that bears the cost and owns the results claims; work done for a client who pays for it is a contract payment to the performer, which the client claims.
Illustrative example
Illustrative example. Three hypothetical companies. Saguenay Analytics does not know whether a graph-based model can detect fraud patterns across 50 million transactions within a two-second window; it tests three architectures against a benchmark and records why each did or did not meet the target. SR&ED, with the benchmarks as evidence; the alerting dashboard is development. Red Deer Fabrication develops a welding procedure for a new high-strength steel, running 40 test coupons through a designed matrix. SR&ED, with the coupons as materials consumed. Okanagan Beverages reformulates a cider to taste drier. Not SR&ED, a style change. Its parallel project to keep carbonation stable for 12 months in a lightweight can, with monthly measurements, is experimental development.
Related pages
Official sources
Educational content only, not tax advice. See the disclaimer.