If you want building permit processing time benchmarks that actually mean something, you need real numbers from comparable cities, not a definition of what “processing time” is. So here is the headline figure: in Vancouver, a New Building permit takes a median of 168 days from application to issuance, with an average of 201.5 days. That is the single longest median in our dataset, and it tells you immediately how wide the gap between Canadian municipalities really is.
These benchmarks come from our own permit dataset: 4,291,388 permits across 35 cities, covering the trailing 12 months as of June 2026. Every figure below is measured application-to-issuance, and we only report cities with at least 50 valid records so the medians are not noise.
Application-to-Issuance Time by City (as of June 2026)
The fastest and slowest cities are not close. Below are the per-city benchmarks, ranked from quickest median to slowest. The median is the more honest benchmark here: averages get dragged up by a handful of very long files, so the median tells you what a typical applicant actually experiences.
| City | Avg days | Median days | Records (n) |
|---|---|---|---|
| Thunder Bay | 29.2 | 10 | 910 |
| Kelowna | 39.8 | 13 | 1,470 |
| Montreal | 50.9 | 20 | 18,381 |
| Toronto | 71.7 | 28 | 33,798 |
| St. Catharines | 114.3 | 62 | 1,475 |
| Vancouver | 112.9 | 71 | 4,297 |
Source: Steller Canadian building-permit dataset, trailing 12 months as of June 2026. Cities with fewer than 50 valid records excluded.
The spread is the story. A typical permit clears in 10 days in Thunder Bay and 71 days in Vancouver, a 7x difference between two Canadian cities running the same basic process. Toronto, despite handling the largest volume in the dataset (33,798 permits), still posts a 28-day median, faster than both Vancouver and St. Catharines. Volume alone does not explain slowness.
Why a Single City-Wide Benchmark Misleads You
A city-wide median is a starting point, not an answer, because work types behave nothing alike. Here is Vancouver broken out by what the permit is actually for:
| Work type | Avg days | Median days | Records (n) |
|---|---|---|---|
| Temporary Building/Structure | 23.8 | 13 | 39 |
| Addition/Alteration | 68.0 | 36 | 2,273 |
| Salvage and Abatement | 82.0 | 38 | 268 |
| Demolition/Deconstruction | 144.0 | 118 | 665 |
| New Building | 201.5 | 168 | 1,051 |
Source: Steller Canadian building-permit dataset, Vancouver, trailing 12 months as of June 2026.
A New Building permit (168-day median) takes more than four times as long as an Addition/Alteration (36-day median) in the exact same city. If you benchmark your New Building queue against a blended city-wide number, you will either congratulate yourself for being “average” while builders wait six months, or panic over a delay that is normal for that work type. Benchmark like against like.
How to Build Your Own Benchmark
You do not need an analytics platform to start. You need three columns: application date, issuance date, and work type. From there:
- Compute the median, not just the average, for each work type. As the Vancouver table shows, averages can sit far above the median when a few complex files run long. The median is what most applicants experience.
- Set a minimum record count. We exclude anything under 50 valid records because small samples produce unstable benchmarks. Apply the same discipline to your own categories.
- Compare against peer cities, not a national ideal. Thunder Bay’s 10-day median and Vancouver’s 71-day median are both “normal” for very different cities. Find municipalities with similar volume and project mix before deciding you are slow.
One caution: these benchmarks measure elapsed time from application to issuance only. They do not measure how often permits are approved versus refused, because that outcome is not tracked in the underlying records. Treat these numbers as a timing benchmark, not an approval-rate benchmark.
What These Benchmarks Tell a Municipality
The practical use is triage. When you can see that your Addition/Alteration files are tracking near a 36-day median but your New Building files are stretching past six months, you know exactly where to put staff and process changes. A blended average hides that signal. Per-work-type, peer-compared benchmarks surface it.
This is exactly the comparison Steller’s municipal permit analytics is built to run continuously: every permit, broken out by work type, benchmarked against comparable Canadian cities, updated as new records come in.
Conclusion
Useful building permit processing time benchmarks are specific, segmented, and grounded in real records, not generic targets. The data is unambiguous: medians range from 10 days in Thunder Bay to 71 days in Vancouver, and within a single city a New Building permit can take over four times longer than an alteration. Start by measuring your own application-to-issuance medians per work type, compare against peer cities, and hold yourself to a minimum sample size before drawing conclusions. If you want this benchmarking done automatically across your permit data, request a demo at getsteller.ca.