Statistics Canada publishes three rates that describe the same group of apprentices at the same moment, and they add up to exactly 100. At the end of the expected length of the program, 19.9% had certified and 30.9% had already left. Put side by side, that is 1.55 apprentices walking away for every one who finishes on schedule.
New registrations in Canadian apprenticeship programs hit 101,541 in 2024, a record since the series began in 2008 and the fourth straight year of growth. That is the number that travelled. The three rates published in the same release travelled much less, and they describe something the registration count cannot: what happens to apprentices once they are in.
The rates come from the Statistics Canada release of 11 December 2025, New registrations, certifications and pathway indicators of registered apprentices in Canada, 2024. The reporting that put faces to those numbers was done by Jessica Wong for CBC News, who interviewed apprentices and trainers about what actually stalls a certification. One of them was Joellah Fletcher, a level 2 carpentry apprentice from outside Stratford, Ontario, who chose a trade over university, took carpentry co-op placements and a dual credit in Grade 12, and researched the path largely on her own because her school had little information to offer. She also spent the first year or so of her apprenticeship not keeping the log book that certification requires. She told CBC she was about a year, a year and a half in “before I even realized I was supposed to have a log book”.
Hold that detail. The paperwork is not a footnote to the data. It is one of the mechanisms by which a person who is doing the work ends up counted in a category other than certified.
What the three apprenticeship rates actually count
Statistics Canada evaluates apprentices at the end of the expected duration of their program. Each one is then in one of three states, and the agency publishes all three:
- Certification rate, 19.9%. Completed the program within the expected duration. Up 1.1 percentage points, a third consecutive year of growth.
- Continuation rate, 49.2%. Still registered at the end of the expected duration, not yet certified. Down 0.8 points.
- Discontinuation rate, 30.9%. Left within the expected duration. Down 0.3 points.
Those add to 100.0. We checked that deliberately, because three percentages from the same release do not always share a base, and comparing two that do not is the fastest way to publish a wrong number. Here they do share one. The same test works on the previous year implied by the published changes: 18.8 plus 50.0 plus 31.2 is also 100.0. So the three can be read against each other directly, and the first comparison worth making is the one the release leaves on the table. Divide 30.9 by 19.9 and you get 1.55. For every apprentice who certifies inside the expected duration, roughly 1.55 have already left inside the same window.
One number the release does publish is worth keeping in view next to it: the certification rate is still below the 20.9% recorded in 2019. Three consecutive years of growth have left a gap of 1.0 percentage point, which is 4.8% of the 2019 level.
Worth noting what the 2024 label means. The release states that most apprenticeship programs run four years, so the people measured at the end of their expected duration in 2024 are, for the most part, people who registered around four years earlier. The label is the year the clock ran out, not the year they signed up.
What happens when the same apprentices get eight years instead of four
Here the picture changes, and this is the part the headline number cannot carry on its own. Statistics Canada does not only measure at the expected duration. It follows the same registration cohort out to one and a half times and twice the expected duration, which for a four year program means six years and eight years. The most recent cohort published at all three windows is the one that registered in 2016: 70,470 apprentices, four year nominal duration.
| Status of the 2016 registration cohort | At 4 years | At 6 years | At 8 years |
|---|---|---|---|
| Certified | 20.9% | 34.6% | 38.9% |
| Still registered, not certified | 45.2% | 25.5% | 18.7% |
| Left the program | 33.8% | 39.9% | 42.5% |
Certification nearly doubles. It goes from 20.9% to 38.9%, a gain of 18.0 percentage points and a multiple of 1.86, for the same cohort followed twice as long. That is not a one off. The 2014 cohort multiplied its certification rate by 1.90 over the same stretch, and the 2015 cohort by 1.81.
The more useful question is where the middle group went. Between year four and year eight, the share still registered fell by 26.5 percentage points. Certification picked up 18.0 of those points and departures picked up 8.7. Of the 2016 apprentices who were unresolved at four years and resolved by eight, 67.9% certified and 32.8% left. Roughly two in three finished, late.
Run the leave to finish ratio again on that cohort and it moves from 1.62 at four years to 1.09 at eight. The headline of this piece is a real measurement of a real window. It is also a window that closes on people who are still working.
Women in the 2016 cohort certified faster, and the gap narrowed later
The same table splits by sex. Women were 10,080 of the 70,470 apprentices in the 2016 cohort, or 14.3%. At four years, 25.2% of them had certified against 20.2% of the men, and 28.8% had left against 34.7%. By eight years the two certification figures were 40.2% and 38.6%, so a 5.0 point advantage had narrowed to 1.6. For this cohort the early difference is in timing more than in final outcome.
Where registrations grew, and where they did not
Growth was not evenly spread across the trades. Registrations rose year over year, or held roughly steady, in 16 of the 24 major trade groups, which means eight of them, a third, fell. Interior finishers led the gains at 27.9%, followed by hairstylists and estheticians at 26.5%, construction workers at 24.2%, and early childhood educators and assistants at 16.7%. At the other end, machinists were down 11.5%, stationary engineers and power plant operators down 9.1%, and landscape and horticulture technicians and specialists down 6.5%. From top to bottom that is a spread of 39.4 percentage points inside a national series that rose 5.9%.
By province, registrations fell in six provinces and two territories, the largest declines being Nova Scotia at 7.4% and Quebec at 4.3%. Every province except Quebec is now above its 2019 registration level.
The employer side of this sits closer to the small business owners who sponsor and train than to any national programme, which is the layer the release does not measure. Emily Arrowsmith, director of research at the Canadian Apprenticeship Forum, told CBC that registrations rising is welcome but incomplete without completion, and pointed to level 1 and 2 apprentices struggling to find sponsoring employers, to apprentices reluctant to take a pay cut for the training blocks that move them up a level, and to a general erosion of employer investment in mentorship. Jeff Sloychuk, president of the Yukon Building Trades Council, made a narrower point in the same piece: trades training teaches the craft, not how to mentor someone else in it.
What the data cannot show
The tables carry rates, not reasons. Nothing in them explains why any individual apprentice left, and nothing licenses a causal reading of the trade or provincial differences. A log book started late, a sponsor who cannot release someone for a training block, a move between provinces: all of those land in the same cell, undistinguished.
The 2024 rates and the 2016 cohort table describe different populations, which is why they sit in separate sections here. The eight year view exists only for cohorts old enough to have been followed that long, so it cannot tell you what the apprentices measured in 2024 will do by 2032.
The release also notes administrative changes that affect comparisons over time. Alberta designated a number of occupations as trades in 2023 and began issuing certificates, and from that year apprentices in certain trades could be reported more than once across related sub-trades. Ontario data were revised to remove cases where an apprentice may have been counted as holding both a non Red Seal and a Red Seal certificate in the same trade.
Frequently asked questions
What is the apprenticeship completion rate in Canada?
Statistics Canada reported a certification rate of 19.9% for 2024, meaning 19.9% of apprentices completed their programs within the expected duration. That was up 1.1 percentage points from 2023 and still below the 20.9% recorded in 2019.
What is the difference between the continuation rate and the discontinuation rate?
The continuation rate, 49.2% in 2024, is the share of apprentices still registered at the end of the expected duration of their program but not yet certified. The discontinuation rate, 30.9%, is the share who left within that same window. Together with the 19.9% certification rate they account for the whole group.
How long does an apprenticeship take in Canada?
Most apprenticeship programs have an expected duration of four years, which is why Statistics Canada also reports outcomes at six years and eight years, that is at one and a half times and twice the expected duration. Individual outcomes are evaluated against each apprentice’s own program length, which can differ from the most common value in their jurisdiction.
Do apprentices who miss the expected duration ever certify?
Many do. Of the apprentices who registered in 2016, 20.9% had certified at four years and 38.9% had certified at eight. Of those unresolved at four years who resolved by eight, Raw POV calculates that 67.9% certified and 32.8% left.
Sources and method
From the source. Registration, certification and rate figures for 2024 come from Statistics Canada, The Daily, New registrations, certifications and pathway indicators of registered apprentices in Canada, 2024, published 11 December 2025. We read the release itself rather than a summary of it, and the definitions of the three rates quoted above are the agency’s own. Cohort figures at four, six and eight years come from Statistics Canada table 37-10-0193-01, which we downloaded in full and read for Canada, all trades. The interviews with Joellah Fletcher, Emily Arrowsmith and Jeff Sloychuk were reported by Jessica Wong for CBC News on 21 December 2025. One short quotation is reproduced here; the rest is paraphrased.
Our own analysis. Statistics Canada publishes none of the following. Raw POV calculated each: the ratio of 1.55 between the 2024 discontinuation and certification rates; the check that the three 2024 rates sum to 100.0, and that the implied 2023 rates do too; the 4.8% relative size of the remaining gap to 2019; the 18.0 point gain and 1.86 multiple for the 2016 cohort between four and eight years, plus multiples of 1.90 and 1.81 for the 2014 and 2015 cohorts; the split of the 26.5 continuation points into 67.9% certified and 32.8% departed; the fall in the leave to finish ratio from 1.62 to 1.09; the narrowing of the certification gap between women and men from 5.0 points to 1.6; the count of eight declining trade groups out of 24; and the 39.4 point spread across trade groups.
Bases, stated plainly. The 2024 rates and the 2016 cohort table describe different people and are never compared as if they were the same. The 2024 label refers to the year an apprentice reached the end of the expected duration of the program, not the year of registration. Registration counts and certification counts for 2024 are flows from different cohorts and are not divided into one another anywhere in this article.
Related reading on credentials and the Canadian labour market: our analysis of where a degree was earned and overqualification, the piece on what Canada’s repair law did and did not oblige, and the evidence on who actually gets trained at work.
