Insights
What Triggers a CRA Audit in 2026
How the CRA actually selects business files in 2026 — the data-matching triggers, the targeted industries, and what still gets picked at random.
The short answer
CRA selects most business tax audits by risk, not chance. The dominant driver in 2026 is data matching: a corporate T2 that does not reconcile with the GST/HST returns, third-party slips (T5018, T4A, and others) that do not match what was filed, industry-benchmark outliers, unusually high input tax credits, repeated losses, and lifestyle that outruns reported income.
What Actually Flags a Return for CRA Review?
Most business tax audits are not random. CRA runs filed returns through risk-assessment systems that score each return for the likelihood of error or non-compliance, and an officer then reviews the higher-scoring files before a CRA audit is opened. The agency describes the selection as risk-based: returns are compared against information from third parties, against the taxpayer's own filing history, and against other businesses in the same industry.
In 2026 the most productive input to that scoring is data that does not reconcile. A corporate income tax return (T2) reporting revenue the GST/HST returns do not, third-party slips that do not match what was filed, input tax credits that are large relative to sales, and margins or expense ratios that sit well outside the industry norm — each is a discrepancy a computer can find at scale, without an auditor first reading a page.
The other recurring flags are behavioural rather than arithmetic: several years of business losses claimed against other income, round-number or steadily growing shareholder loan balances, a sudden change in a major figure, and reported income that does not support the assets and spending CRA can observe. None of these is proof of anything. Each simply raises the probability that a review finds something — which is exactly what a risk score measures.
- A T2 corporate return whose revenue does not reconcile to the GST/HST returns for the same period — the top data-driven trigger
- T4, T5, T5018, or T4A third-party slips that do not match the income reported on the return they belong to
- Input tax credits that are unusually high relative to reported sales on an HST/GST return
- Gross margins or expense ratios well outside the benchmark for the business's industry
- Repeated business losses claimed against other income year after year
- Large, round, or steadily rising shareholder loan and intercompany balances
- Reported income that cannot support observable assets, deposits, or lifestyle
What Is a Data-Matching or Third-Party Trigger?
A data-matching trigger is a mismatch between two records CRA already holds. Each year CRA receives hundreds of millions of third-party information slips — T4s from employers, T5s from banks, T5008s from brokers, T5018s from construction contractors, T4As for fees for service — and matches them, by social insurance or business number, against the returns the recipients file. When a slip reports income a return does not, the system flags it. The taxpayer never sees the match happen; the letter that follows is its output.
The most powerful internal match is between a corporation's income tax return and its own GST/HST returns. The two are filed separately, often by different people, and CRA reconciles the revenue reported on the T2 against the sales reported for GST/HST over the same period. A gap in either direction — HST collected on sales that never reached the income statement, or income reported without the matching taxable supplies — is a clean, automatable signal that something was not filed consistently. It is, in practice, the top data-driven trigger for a business tax audit.
Third-party reporting is expanding, not contracting. In construction, the Contract Payment Reporting System requires a contractor to file a T5018 for every subcontractor paid more than $500 in a year, and those payments become income CRA expects on the subcontractor's return. In trucking, the T4A moratorium — a temporary pause on penalties for unreported fees for service — was lifted for the 2025 tax year, so payments to incorporated drivers are now reported and matched like any other slip. Each new reporting stream gives CRA another return to reconcile.
Which Industries Does CRA Target in 2026?
CRA does not audit every sector at the same rate. It concentrates on industries where the tax gap is documented and where cash, subcontracting, or worker misclassification make under-reporting easier. Construction, trucking, food services, and real estate recur in the agency's own compliance reporting, and the tools differ by sector.
In construction, the T5018 system feeds third-party matching: the subcontractor payments a general contractor reports become income CRA expects to see filed. In trucking, the focus is the “Driver Inc.” model — drivers incorporated to receive fees without source deductions — and with the T4A moratorium lifted for 2025 and Budget 2025 committing roughly $77 million to enforcement, those arrangements now generate matched slips and personal-services-business scrutiny.
In food services and other cash-intensive businesses, CRA runs point-of-sale audits and uses dedicated teams to detect electronic sales suppression (“zapper”) software that deletes transactions; where records are unreliable, it verifies income indirectly, reconstructing sales from purchases, deposits, and spending rather than from the books. In real estate, the flipping rule applies: since January 1, 2023, a residential property held under 365 days is taxed as business income with no principal residence exemption, and CRA matches property dispositions against reported gains. Across its 2015–2018 underground-economy effort, CRA reported more than $4 billion in unreported income from over 17,000 audits, much of it through these point-of-sale and real-estate teams.
Do Random Audits Still Happen?
Yes, but they are a small share of the total. CRA lists random selection as one of several reasons a file is chosen, and it does run some genuinely random audits — partly to test compliance across a population, and partly to calibrate the risk models that drive everything else. For most business owners, the odds of a purely random audit in any given year are low.
The practical consequence is that “I did nothing to attract attention” is a weak position, because most attention is not attracted at all — it is computed. A return can be selected because a slip did not match, because a ratio was an outlier, or because the industry itself carries a higher audit rate, none of which requires an auditor to have suspected the taxpayer of anything. Understanding that the trigger is usually data, not judgment, changes how a file is kept: the aim is reconciliation that survives a match, not the avoidance of some list.
It also means a review is not an accusation. A large share of tax audits begin as verification of a single discrepancy and close once it is explained. What decides the outcome is whether the records answer the question the data raised — which is settled long before the letter arrives.
How Do I Lower My Audit Risk Without Red Flags?
The most effective step is also the least visible: file consistently across returns. Reconcile the revenue on the T2 to the sales on the GST/HST returns before filing, confirm that every third-party slip received is reflected on the return it belongs to, and resolve differences rather than leaving them for CRA to find. Many data-matching flags are not evidence of under-reporting at all — they are timing differences and posting errors that were never reconciled.
On the input tax credit side, keep the documentation CRA actually requires. The agency sets it out by dollar band — minimal information under $30, the supplier's GST/HST registration number from $30, and full particulars at $150 and over — and it expects a claimant to show the supplier was registered on the date of the supply. Claiming credits that are large relative to sales without that paper is a common and avoidable HST/GST audit trigger.
Two structural items deserve standing attention. Shareholder loans and intercompany balances should be documented, supported by agreements, and cleared within the timelines the Income Tax Act allows, because large or round balances draw questions. And where cash or personal spending could outrun reported income, the answer is a contemporaneous record of non-taxable sources — loans, gifts, asset sales, savings — so an indirect review has something to reconcile against. None of this is about looking innocent; it is about being able to answer, on paper, the questions the data will eventually ask.
By the numbers
The figures behind this
- the amount above which a construction contractor must report each subcontractor's payments on a T5018 — third-party data the CRA matches against the subcontractor's own return
- $500
- in unreported income CRA identified through more than 17,000 underground-economy income tax and GST/HST audits from 2015 to 2018, led by its point-of-sale (restaurant and bar) and real-estate teams
- $4 billion
- hold a residential property for fewer than this and, since January 1, 2023, the gain is fully taxable business income with the principal residence exemption denied (Income Tax Act s.12(12))
- 365 days
the amount above which a construction contractor must report each subcontractor's payments on a T5018 — third-party data the CRA matches against the subcontractor's own return
Source ↗in unreported income CRA identified through more than 17,000 underground-economy income tax and GST/HST audits from 2015 to 2018, led by its point-of-sale (restaurant and bar) and real-estate teams
Source ↗hold a residential property for fewer than this and, since January 1, 2023, the gain is fully taxable business income with the principal residence exemption denied (Income Tax Act s.12(12))
Source ↗“A CRA audit rarely begins with suspicion — it begins with two numbers that were meant to agree and didn't, and a file is only ever as safe as the reconciliation no one troubled to keep.”
Frequent questions
Questions this raises
Does filing my GST/HST returns and my T2 separately stop CRA from comparing them?
- No. The two are filed on different schedules and often by different people, but CRA reconciles them by business number. It compares the revenue reported on the corporate income tax return against the sales reported for GST/HST over the same period, and a gap in either direction is one of the most common data-driven audit triggers. Reconciling the two before filing — and documenting any legitimate difference — is the single most useful habit for a GST/HST-registered business.
Are shareholder loans really an audit flag?
- Large, round, or steadily growing shareholder loan and intercompany balances draw CRA's attention because they can mask unreported benefits or income. They are not improper in themselves. What matters is that each is documented, supported by an agreement, and cleared within the timelines the Income Tax Act allows. A balance that is explained and repaid on schedule is routine; one that grows without records invites questions.
Will claiming a business loss for several years trigger a CRA audit?
- Repeated losses claimed against other income are a recognized risk factor, not an automatic CRA audit. The agency distinguishes a genuine business going through hard years from an activity with no realistic path to profit, which it may treat as a non-deductible personal pursuit. The protection is evidence of commercial intent — a business plan, real marketing, arm's-length pricing, and a credible route to profit — kept contemporaneously rather than assembled after a letter arrives.
Do unusually high input tax credits get a business audited?
- Input tax credits that are large relative to reported sales are a frequent HST/GST audit trigger, because they can indicate either legitimate investment or over-claimed and unsupported credits. The determining factor is documentation. CRA sets out what a claim must show by dollar band and expects proof that each supplier was registered for GST/HST on the date of the supply. Credits backed by that record withstand review; credits without it are commonly denied.
Can CRA compare my lifestyle to the income I report?
- Yes. Where reported income does not appear to support observable assets, deposits, or spending, CRA can verify income indirectly — the net worth or deposit-analysis method — reconstructing income from the outside rather than from the books. Because an assessment is presumed valid once issued, the taxpayer then carries the burden of showing the numbers are wrong. A contemporaneous record of non-taxable sources such as loans, gifts, and asset sales is what makes that burden answerable.
References
Primary sources
- CRA — Business audits: how the CRA selects files
- CRA — What you should know about audits (RC4188)
- CRA — T5018 Statement of Contract Payments (Contract Payment Reporting System)
- CRA — Strengthening compliance in trucking: lifting the T4A moratorium (Dec 2025)
- CRA — GST/HST Memorandum 8-4: Documentary requirements for claiming input tax credits
- Income Tax Act, s.12(12)–(13) — flipped (residential) property rule
- CRA — Results of the 2015–2018 Underground Economy Strategy
- CRA — Electronic suppression of sales (zapper software) tax alert
Written by
FCCA (United Kingdom)
Published July 10, 2026 · Updated July 10, 2026 · 9 min read
This article reflects tax law and CRA administrative practice as of July 10, 2026. It is general information, not tax, accounting, or legal advice, and reading it does not create a professional-client relationship. Figures, deadlines, and administrative positions change — obtain advice on your own facts before acting.
When the question stops being general.
This analysis sets out how the issue works in principle. A live file turns on its own facts, correspondence, and deadlines. The advisory page below explains how a matter like it is handled in practice; a private consultation is the direct route for a specific situation.
Private consultations available by request. Personally answered — typically within business hours. WhatsApp: +1-647-510-8878.