HSA Eligibility Rules for Shareholder-Employees in Canada

For a shareholder-employee of a Canadian corporation, an HSA is only tax-free if it qualifies as a Private Health Services Plan under the Income Tax Act. That single test, buried in section 248(1), decides whether every dollar reimbursed comes out clean or lands back on a T4 as taxable income. Getting it wrong doesn't just cost the employee: it strips the corporation of its deduction too.

The Income Tax Act never actually defines "Health Spending Account."" What it defines is a Private Health Services Plan, or PHSP, and an HSA only gets its tax treatment by fitting inside that box. Section 248(1) describes a PHSP as a contract of insurance for hospital or medical expenses, or a medical and hospital care insurance plan, explicitly carving out government health plans. CRA's Interpretation Bulletin IT-339R2 and Income Tax Folio S2-F1-C1 fill in the details, and paragraph 3 of IT-339R2 lays out what's known as the "insurance in nature" test: the plan has to work like real insurance, with one party genuinely indemnifying another against an uncertain loss and taking on real risk in doing so. On top of that, CRA requires that all or substantially all premiums, generally read as 90% or more, go toward expenses that qualify for the Medical Expense Tax Credit.

Pass that test and the reward is significant. Reimbursements aren't taxable to the employee, there's no T4 inclusion, no provincial tax exposure (Quebec is the exception; Revenu Québec still taxes these amounts even though they're exempt federally), and the corporation deducts the full cost as a business expense. Fail it, and the whole arrangement flips: the benefit becomes ordinary taxable income to whoever received it, and the corporation loses the deduction it claimed.

Every rule that follows, for the size of the plan, who can be on it, how it's administered, traces back to that one question: does this arrangement actually look like insurance between two separate parties? For a shareholder-employee, especially one who owns the whole company, that question gets a lot harder to answer than it looks.

How incorporation changes the eligibility picture entirely

Whether a business is incorporated changes everything about HSA eligibility, right from the start.

Any incorporated business can offer an HSA, whether it's a multinational with thousands of staff or a one-person professional corporation. Incorporation creates a separate legal entity, and that separation is what lets the owner also be a legitimate employee of the company. CRA accepts this arrangement, per its Employer's Guide to Taxable Benefits and Allowances (T4130), as long as the corporation pays the owner a regular salary and issues a T4.

Sole proprietors face a completely different rule. An unincorporated business only qualifies for HSA treatment if it employs at least one arm's-length worker, someone who isn't a family member or the owner. No arm's-length employee means CRA treats any health-spending-type benefit as personal income, not a deductible business expense. Statutory caps under section 20.01(2) of the Income Tax Act apply to unincorporated sole proprietors specifically (that section governs the dollar limits, while IT-339R2 still governs whether the plan qualifies as a PHSP in the first place).

So incorporation isn't just about liability protection or how a business raises capital. For an owner who wants HSA access, it's the entry ticket. Without it, the door doesn't open.

But incorporating doesn't automatically clear the way. Once a business is incorporated, the shareholder-employee still has to clear three more hurdles before any HSA dollar is safe from tax.

The three conditions a shareholder-employee must satisfy to receive HSA benefits tax-free

Condition 1: T4 employment income. The shareholder needs to actually be paid a salary, reported on a T4, rather than taking compensation purely through dividends. This distinction matters more than it sounds. Dividends mark someone as a shareholder in CRA's eyes, not an employee, and an HSA is by definition an employee benefit. If CRA reviews a company that's deducted PHSP costs for someone with zero T4 income, it can reclassify the entire thing as a shareholder benefit under section 15(1). The consequence is blunt: the full value gets added to the shareholder's income as regular income, and the corporation's deduction disappears.

Condition 2: Active engagement in the business. CRA's Technical Interpretation 2003-0050541E5 says the shareholder has to be genuinely involved in day-to-day operations, not just holding shares and collecting a benefit. Incorporation alone doesn't satisfy this. The HSA needs to flow from the person's actual job, and a written employment contract that spells out the role, the responsibilities, and the HSA as a term of that employment goes a long way toward proving it.

Condition 3: Benefits tied to employment, not ownership. The HSA has to reflect what someone earns as an employee, never as a shareholder. In practice this means a handful of things line up together: the plan goes to employees, not to shareholders as such; it treats shareholder and non-shareholder employees the same way; and people in similar roles get similar treatment. If CRA concludes the benefit really flowed from share ownership, the whole thing gets taxed in the shareholder's hands, and the corporation loses its deduction along with it.

Why the sole shareholder-employee is the hardest case, and what CRA has actually said

All three conditions get dramatically harder to satisfy when there's only one person in the company. That's the sole shareholder-employee problem, and it sits at the center of nearly every HSA dispute involving small, owner-only corporations.

Once spelled out, the structural issue is straightforward: insurance requires two parties, one bearing risk on behalf of another. When the shareholder and the employee are the same person, there's no independent party left to actually assume that risk. The whole "insurance in nature" test starts to wobble.

CRA addressed this directly in Technical Interpretation 2014-0521301E5, stating that a plan for a sole employee-shareholder would likely not qualify as a PHSP, since it lacks the necessary elements of insurance. But CRA also offered a path forward, of sorts. Its stated position: a sole shareholder-employee is presumed to receive the benefit as a shareholder, unless that person can show that non-shareholder employees, doing similar work at a similarly sized company, receive similar benefits under a similar plan. In other words, the burden of proof sits entirely with the shareholder-employee.

The 2022 CALU Roundtable (Question 10, May 3, 2022) sharpened this further, and it's the most recent authoritative word CRA has given on the topic. The scenario: a cost-plus plan where an administrator reimburses a sole employee-shareholder, plus their spouse and household, for actual medical expenses, and collects the reimbursed amount plus a fee in return. CRA's answer was that this does not qualify as a PHSP. There's no genuine risk being assumed by anyone. The sole employee-shareholder is, functionally, just running personal medical bills through the corporation. CRA's own language on this bears sitting with: if there's little chance the employee won't eventually get reimbursed for the full amount allocated each year, the arrangement isn't insurance, full stop.

And the exposure doesn't stop at the PHSP question, either. Even in a scenario where the insurance-in-nature test were somehow met, section 15(1) still applies if the benefit was received in the person's capacity as shareholder. Two separate doors, either one of which can lead to the same tax bill.

Some HSA providers in the market argue there's a workaround: amend the employment contract so the HSA becomes a binding obligation the corporation cannot modify or cancel unilaterally, which introduces the contractual risk CRA is looking for. That argument has some logic to it, and it's a legitimate structuring option worth discussing with a tax advisor. But it hasn't been blessed by CRA in the CALU statement, and providers who suggest it generally recommend running it past a tax professional first, for good reason.

The gap between what CRA has said in writing and what gets marketed to sole shareholder-employees is wide. That gap is exactly where compliance risk lives. Nobody should treat this as a settled question just because a provider is willing to set up the plan.

How to document an HSA so it withstands CRA scrutiny

Paperwork does a lot of the heavy lifting here, more than most owners expect.

Start with a written plan document. It needs to name the eligible employees, set annual limits by employee class, list eligible expenses, describe the claims process, and say clearly whether unused funds roll over or get forfeited. Limits can't be unlimited for any class of employee: they need to be reasonable and applied the same way across everyone in that class. The plan also has to be funded entirely by the employer; employees can't put their own money in. Every T4 employee in a given class has to be included in that class (part-timers can be split into their own class or left out, but whatever category they're in has to be applied consistently), and contractors who don't get a T4 don't qualify at all. Employees also need to actually know the plan exists and be able to see its terms.

For a sole shareholder situation specifically, the corporate paperwork matters even more. An employment contract should name the HSA as a term of employment, specify the limits and coverage, and state outright that the corporation can't modify or cancel the plan on its own or at the shareholder's whim. It's also worth having the corporate minute book note that equivalent benefits would be extended to future arm's-length hires. That single note supports the comparator argument CRA explicitly asked for in its 2014 interpretation.

Third-party administration is not optional. A corporation cannot review and approve its own claims: Having an independent administrator sitting between the plan and the person submitting expenses is strongly recommended for compliance purposes. This matters more than it might seem. A shareholder-owner signing off on their own reimbursements undercuts the insurance-in-nature test almost by itself, since there's no arm's-length check on anything. A genuine third-party administrator reviews claims against CRA's eligible expense list, keeps a paper trail, and provides exactly the adjudication that supports PHSP status. Administration fees for this service typically run in the high single digits to low double digits as a percentage of claims, and those fees are themselves deductible. For smaller incorporated businesses, pay-as-you-go providers exist that charge on a per-claim basis, which keeps this structure workable without a large upfront outlay.

Finally, benchmark against comparable roles. CRA expects an HSA allocation for a shareholder-employee to look like what an arm's-length employee doing similar work at a similarly sized business would get. Documenting that comparison, even informally, strengthens the case that the benefit is coming from the job, not the ownership stake.

What "reasonable" means in practice for annual HSA allocations

There's no fixed dollar ceiling on contributions the way there is with certain other registered savings vehicles. CRA doesn't publish a cap. Instead, it applies a reasonableness test: the allocation has to make sense as compensation for the work being done, not as a way to funnel personal medical costs through the corporation tax-free.

Providers in the industry have developed their own rules of thumb, and they're worth knowing, though none of them are CRA law. Some cap annual HSA credits around 20% of an employee's gross T4 salary, so someone earning $60,000 might have a maximum of near $12,000 a year in HSA credits. Others cite 25% of gross salary as a broader outer limit used across the industry, while some providers work off a tighter range, closer to 10 to 15% of what a fairly paid employee in the same role would earn. At least one provider guideline puts an absolute dollar ceiling around $25,000 a year, regardless of salary.

None of these numbers come from CRA directly. They're safe harbors that providers have built based on what tends to survive scrutiny, and what CRA actually cares about is comparability.

The warning sign is obvious once it's named: a sole shareholder-employee allocating far more than a market-rate employee in the same role would ever receive. That gap itself becomes evidence, and not the kind anyone wants sitting in a CRA file, that the benefit is coming from ownership rather than employment.

There's a practical wrinkle to name too. Owners who keep their T4 salary low, a common move to reduce payroll tax, run into a problem if HSA limits are set as a percentage of that salary: the HSA ceiling shrinks right along with it. Trying to minimize salary while maximizing HSA dollars creates exactly the kind of inconsistency that draws attention. And whatever number gets settled on, it has to be applied evenly. Employees in similar roles, with similar tenure, need the same annual credit. Giving the shareholder-employee a materially richer benefit than everyone else on the team is its own compliance risk, separate from the dollar amount itself.

The difference between a compliant HSA benefit and a taxable shareholder benefit, and how CRA draws the line

Section 15(1) of the Income Tax Act is the mechanism that makes all of this enforceable. It says that when a corporation gives a benefit to a shareholder, the value of that benefit gets added to the shareholder's income as regular income, and the corporation's deduction disappears in the same stroke.

Whether a benefit counts as flowing from "capacity as shareholder" is a factual question, and CRA weighs the evidence rather than applying a bright-line rule. The factors it looks at read like a checklist: is there T4 income, and is it reasonable for the role? Is the shareholder actually working in the business day to day? Does the HSA allocation match what an arm's-length employee in a similar position would get? Are all T4 employees included in the plan, or just the owner? Is an independent administrator actually reviewing claims? Does the plan document stop the shareholder-employee from changing or canceling the plan whenever it's convenient?

Line those factors up one way, and the case builds itself. There is T4 income, active day-to-day involvement, a formal employment contract naming the HSA, an independent administrator, and allocations that track market pay. That's a shareholder-employee who looks, on paper and in practice, like an employee first.

Line them up the other way, dividends only, allocations well above any employment comparator, no independent claims review, no written plan, and the opposite case builds itself just as easily. In that scenario, the full reimbursement history becomes taxable, and there's little room left to argue otherwise.

CRA's 2022 CALU statement remains the clearest signal of where the agency stands today, and it leaves the sole shareholder-employee scenario as the single riskiest structure in this entire area of tax law. Nobody should set one up without qualified tax advice specific to their situation. Corporations with multiple employees sit in a far more comfortable position: the employment relationship is easier to prove, there are real comparators inside the plan itself, and the insurance-in-nature test has actual people and actual risk behind it, rather than one person writing checks to themselves through a corporate structure.