Is Your Trades Business Actually Eligible for SR&ED?

Most Ontario trades businesses that ask about SR&ED are not eligible, and the honest thing to do is say that plainly before explaining the details. SR&ED (Scientific Research and Experimental Development) is the federal government’s largest R&D tax incentive by dollar volume, and it is real, well-funded, and administered continuously by the Canada Revenue Agency. It is also one of the most misunderstood programmes in Canadian small business, and the misunderstanding runs in a predictable direction: businesses assume that “we did something technically hard” is the bar, when CRA’s actual bar is narrower and more specific than that.

This page walks through the real eligibility test, drawn directly from CRA’s own SR&ED eligibility guidance, checked 2026-08-28, with concrete trades examples of what clears the bar and what doesn’t.

The three-part test CRA actually applies

CRA’s own eligibility page sets out three criteria that work together, not a checklist where any one is enough on its own:

1. Technological uncertainty. The work has to address a situation where it is genuinely unknown, based on existing knowledge, whether a given result can be achieved, or how to achieve it. This is not “we weren’t sure it would work the first time we tried it.” It means the answer wasn’t knowable in advance from published science, standard engineering practice, or your own professional experience, and had to be found out through investigation.

2. Systematic investigation. The work has to follow a structured process CRA describes in four steps: defining a problem, forming a hypothesis about how to resolve it, planning and testing that hypothesis through experiment or analysis, and drawing conclusions from the results. This is closer to how an engineering R&D team documents its work than how a job site normally runs. Notably, and usefully, the work still counts even if the hypothesis turns out to be wrong. Generating new understanding through a failed experiment is still eligible; it’s the process that has to be systematic, not the outcome that has to be successful.

3. Technological advancement. The work has to generate new knowledge that moves scientific or technological understanding forward, even in a narrow, specific way. It does not need to be a breakthrough. It does need to be new, in the sense that the knowledge didn’t already exist and wasn’t readily available before the work was done.

All three have to be true at once. Work that’s difficult but doesn’t involve genuine uncertainty about the outcome fails the first test. Work that involves real uncertainty but was resolved by trial and error without a documented, structured process fails the second. Work that was systematic and uncertain but only produced a result useful to one business, without adding to the underlying technical knowledge, fails the third.

What CRA explicitly says does not qualify

CRA’s own guidance names specific categories of work that are not eligible, regardless of how skilled or valuable the work is:

• Routine engineering or standard practice, including applying known techniques or existing knowledge to a new situation
• Routine quality control and testing
• Market research and sales promotion
• Style changes, and routine data collection
• Commercial production and normal business operations
• Training staff or on-the-job learning
• Hiring an expert to apply knowledge that already exists, rather than to develop new knowledge
• Buying proprietary knowledge or technology that already exists elsewhere

Read that list against a typical trades business. Installing a manufacturer’s system according to its documented specifications: routine practice. Troubleshooting a fault using standard diagnostic procedure, even a genuinely hard one: applying existing knowledge. Training a technician on a new tool: explicitly excluded. Building a piece of internal software by wiring together a database, a scheduling library, and a payment processor’s API, however well done: usually a combination of known techniques, not a technological advancement in itself.

Where the line genuinely sits, with trades-relevant examples

Not eligible, almost always:

• Installing HVAC, fire-safety, electrical, or plumbing equipment to spec, even complex jobs, even equipment that is new to the Canadian market
• Diagnosing and repairing equipment faults using known diagnostic methods, however hard-won that skill is
• Adopting new field-service or scheduling software, even software built specifically for your business, if it combines existing, well-understood components in a new arrangement
• Training staff on new tools, codes, or procedures
• Modifying a standard installation to fit an unusual building, if the modification draws on known trade knowledge rather than resolving a genuine unknown

Potentially eligible, on a case-by-case basis, and only with real technical uncertainty and documentation:

• Developing a genuinely new method, material, or process where the outcome could not be predicted from existing engineering knowledge, and testing it through a documented, structured process
• Building software or a control system that has to solve a problem existing tools and known techniques cannot solve, where the development itself required systematic experimentation to find out whether and how it could be done
• Adapting a technology to a use case so far outside its documented specifications that applying it required genuinely new technical knowledge, not just trial and error within known parameters

The second list is narrow on purpose. It is where a small number of trades and adjacent businesses, usually ones doing genuine product development, custom equipment design, or first-of-its-kind process engineering, can have a real claim. It is not where most day-to-day trades work sits, no matter how skilled that work is.

Why this matters before you pay anyone

SR&ED claim preparation is itself a paid service, and some firms work on a contingency basis against the eventual credit. That fee structure means a firm has a financial incentive to encourage a claim even when the underlying work is a poor fit for the criteria above, because their downside if the claim is weak is limited and their upside if it’s approved is real. That isn’t an accusation against any specific firm; it’s a structural incentive worth knowing about before signing anything. Read the three-part test above, honestly, against your own business first. If nothing on your job sites in the last year involved genuine, undocumented-in-advance technical uncertainty resolved through a structured process, the honest answer is that SR&ED probably isn’t for you this year, and no consultant’s enthusiasm changes that.

If, after reading this, your business genuinely has a project that looks like the second list, the next step is a conversation with an accountant or SR&ED preparer who works with CRA’s criteria regularly, not a general guide like this one. And before that conversation, read the companion guide, What a Grant Application Actually Requires, because SR&ED’s documentation burden is real, and it’s better to know the size of that job before you start it than to find out partway through.