Archived edition · View the latest edition
The autumn comparison depends on the property type. A single statement that the market sped up would miss the different experiences across these four categories.
Selling times by property type
| Property type | Sept. 2025 median days | Aug. 2026 median days | Sept. 2026 median days | Sept. middle half, days | Sept. valid episodes |
|---|---|---|---|---|---|
| Single detached | 17.5 | 25.5 | 21 | 11.75–45 | 388 |
| Freehold townhouse | 20 | 26 | 26 | 15–46 | 225 |
| Apartment condominium | 37 | 32.5 | 30 | 18–53 | 109 |
| Townhouse condominium | 26 | 33 | 34.5 | 21–60 | 80 |
The middle-half range runs from the 25th to the 75th percentile. It describes the spread of recorded results, rather than a deadline or forecast for a new listing. Half-day medians arise when the two middle observations are averaged.
Detached homes had a median of 21 days, down 4.5 days from August's 25.5. Apartment condos moved from 32.5 to 30 days. Freehold townhouses stayed at 26 days, while condo townhouses moved from 33 to 34.5 days.
Does the annual comparison tell the same story?
No. Detached homes were faster than August but slower than September last year: their median increased from 17.5 to 21 days. Freehold townhouses increased from 20 to 26 days, and condo townhouses from 26 to 34.5. Apartment condos were the exception, falling from 37 to 30 days.
Those are descriptive comparisons of different groups of sold homes. They are not seasonally adjusted, and they do not isolate a change in demand from changes in the homes that sold.
What should a seller do with this information?
Start with the relevant category, then examine comparable homes and competing listings in the immediate area. The detached category's September middle half ran from 11.75 to 45 days; that spread is a reason to discuss a range of outcomes rather than promise the median.
Selling speed also does not establish how close the price was to asking. The separate below-asking comparison measures that question directly.
A measurement change from the original August article
The original August edition used whole-day figures from supplied PropTx-based reports. This edition calculates medians directly from the feed's DaysOnMarket field for all three comparison months. The provider defines that field by MLS business rules but has not supplied the full day-count or relisting conventions in the dictionary.
We therefore do not splice the old report figures into this comparison, add a day, or treat a difference between the two methods as a market change. The August archive remains a record of the original report-based edition.
How to use the comparison in a selling plan
Start with the home type and the range around the median, then compare recent sales and competing listings in the relevant area. For a review with Matt, bring the property details, likely selling window and any deadline that matters. The local evidence can inform that discussion; these sold-only medians cannot promise a sale date.
Source and methods
Data source: PropTx, retrieved October 6, 2026, beginning at 23:53 UTC (7:53 p.m. Toronto). New Purveyors Lab calculations use records marked Sold and For Sale, with CountyOrParish equal to Ottawa. This provider geography is not asserted to match the municipal boundary or the entire Ottawa board territory. Reporting months use PurchaseContractDate—the recorded date of the accepted sale agreement—not the closing date.
The September analysis includes 802 sold listing episodes: 388 single detached, 225 freehold townhouses, 109 apartment condominiums and 80 townhouse condominiums. Another 85 records in the broad residential extraction fall outside those four categories. The comparison samples contain 788 selected episodes in August 2026 and 893 in September 2025. Each headline metric has its own valid denominator; all selected records in these three months have usable sale prices, both asking-price fields and nonnegative provider days on market.
All three periods were retrieved in the same October 6 run. Two August episodes present in the September 19 extraction are no longer returned; the shared price and duration fields of retained records are unchanged. This is a source revision, not market activity occurring between the extraction dates. Prior published editions retain their original figures.
Medians and middle-half ranges use unrounded inputs and linearly interpolated quartiles. At least 30 valid observations are required for a displayed median or percentage denominator. No outlier trimming was applied. Relistings are not linked into unique homes. Withheld records, later updates, changing sales mix and unmeasured location, size and condition differences limit interpretation. These sold-only samples do not measure available inventory or the probability of selling.
Discuss these findings
For help applying these findings to a property or a move, contact Matt Richling. Include the area, property type and timing that matter to you.
Cite this edition
Richling, Matt. “How long did Ottawa-region homes take to sell in September 2026?” New Purveyors Lab. September 2026. Updated October 7, 2026. Permanent edition.