Three job titles, three different pay scales, and a government classification system that only cleanly recognises one of them. That is the state of data salary information in Canada, and it explains why the numbers you find never quite agree.
This guide covers what data scientists, data analysts and data engineers earn here, where the published figures come from, and why the gaps between them are wider than for almost any other technology role. If you are researching a data scientist salary in Canada and finding numbers that range across sixty thousand dollars, the reason is methodological rather than any of them being wrong.
Data scientist salary in Canada
The Government of Canada’s Job Bank classifies data scientists under NOC 21211 and reports a national range of $30.00 to $69.74 an hour, updated November 2025. At full-time hours that is roughly $62,400 to $145,000.
Ontario runs slightly higher at $31.25 to $71.79, and the Toronto region higher again at $32.00 to $71.79, or about $66,600 to $149,300 a year.
The commercial sources land differently:
- Robert Half, Toronto: $108,733 to $163,490. Drawn from their own placement data, which skews toward larger employers paying at the upper end.
- PayScale, Toronto: $91,541. Self-reported base salary, excluding bonus.
- PayScale, Canada: $88,825. The national self-reported average.
So the reported data scientist salary depends entirely on who is counting. Robert Half’s floor sits above Job Bank’s midpoint. Both are accurate for what they measure. Job Bank captures everyone in the occupation including public sector, academia and smaller employers. Robert Half captures the roles a specialist agency gets asked to fill, which are not a representative sample of the market.
Data analyst salary in Canada
Here the classification starts to break down. Job Bank has no code for the modern data analyst role. The nearest match is NOC 21223, database analysts and data administrators, which reports $25.00 to $61.03 an hour nationally and $24.58 to $61.54 in Ontario.
That occupation was defined around database administration, and it captures a meaningfully different job from what most Canadian employers now advertise as a data analyst. Treat those figures as a floor rather than a benchmark.
The commercial sources are closer together on this role than on any other:
- Indeed, Toronto: $78,639 from 212 reported salaries.
- Indeed, Canada: $76,681 from 578 reported salaries.
- Glassdoor, Canada: $70,235.
- Glassdoor, Toronto: $69,944.
- PayScale, Toronto: $67,602.
A spread of about $11,000 across five sources, which is tight by the standards of this field. The reason is volume: analyst roles are common enough that self-reported datasets have real sample sizes behind them.
Data engineer salary in Canada
Job Bank has no code for this role at all. Data engineers get absorbed into software engineering or database administration depending on how a given employer describes the work, which means there is no official Canadian wage report for the title.
In practice the market prices data engineers between analysts and scientists, and often above scientists at senior level. Pipeline and infrastructure work is harder to hire for than modelling work, because the pool is smaller and the skills are less taught. A senior data engineer in Toronto competes with backend and platform engineering salaries rather than with analytics salaries.
If you are benchmarking this role, software engineering data is a closer comparator than data science data. It is also why data engineer searches run differently from the other two.
Why the three roles get confused, and what it costs
The titles describe different work, and our guide to data scientist vs data analyst sets out which role you need, and sorting that out is the first conversation in any data and AI recruitment search. An analyst answers questions using data that already exists. An engineer builds and maintains the systems that make the data usable. A scientist builds models that produce predictions or recommendations.
Canadian employers use them loosely. We regularly see roles advertised as data scientist that are analyst work in substance, and roles advertised as data analyst that are engineering work with SQL in the job title. The aggregators cannot correct for this, because they bucket by the title as posted.
That has a practical cost on both sides. A candidate benchmarking against the wrong title will anchor thousands away from the real market. An employer posting the wrong title attracts the wrong shortlist and then wonders why nobody fits.
The fastest way to sort this out is to describe the first six months of the job rather than the title. We do this at the start of every data scientist and data analyst search. If the answer is dashboards and stakeholder questions, that is an analyst. If it is pipelines and reliability, that is an engineer. If it is models in production, that is a scientist.
What the outlook data says, and it is not what you would expect
Job Bank rates the 2025 to 2027 employment outlook in Ontario as limited for data scientists and moderate for database analysts and data administrators.
The higher-profile title has the weaker outlook. Employment for data scientists is expected to remain relatively stable, with few positions opening through retirement, and roughly a third of them work in professional and technical services with another quarter in finance and insurance.
That fits what we see in the market. Demand for data work is growing, but a great deal of it is being met by tooling, by managed services, or by adding analytics duties to existing engineering roles rather than by hiring dedicated data scientists. Analyst and engineering roles have held up better because they map to ongoing operational work rather than to project-based modelling.
For candidates, the implication is that specialism pays better than generalism right now. For employers, it means the shortage is narrower than the headlines suggest.
What a data hire costs an employer
Every figure above answers the candidate’s question. If you are hiring, the salary is not the number you need to budget.
Add employer costs on top: CPP and EI contributions, employer health tax where it applies, benefits, and equipment. For data roles that usually means cloud and tooling licences as well. A $95,000 analyst is realistically a $115,000 to $125,000 line item before the cost of the search itself, and our 2026 Canadian salary guide covers the equivalent figures across other technology roles.
There is a compliance dimension now too. Since Ontario’s job posting requirements took effect on 1 January 2026, employers with 25 or more staff must publish expected compensation or a range on publicly advertised postings, and a range cannot span more than $50,000. That obligation falls away above $200,000, which covers very few data roles outside leadership.
Because ranges are now public and directly comparable, posting low to preserve negotiating room costs more than it used to.
Using these numbers
A data scientist salary in Canada is not one number, and neither is an analyst’s or an engineer’s. If you are benchmarking an offer, start with Job Bank as your floor. Our breakdown of why salary sources disagree covers what each one measures and how to set a range you can defend, then check whether the commercial figure you are comparing against includes bonus, equity or neither. If a number looks unusually high, it usually includes something the others exclude.
If you are setting a range to post, work from what you can fund and from the seniority you genuinely need, rather than from a national average that blends three different jobs.
If you are structuring bands rather than pricing one offer, our explanation of salary ranges and bands covers how the two differ.