Last reviewed: 28 October 2026. This page is an analysis of published HM Land Registry data. It is general information, not a valuation or financial advice.
About four in ten flats in the London post town (not the outer boroughs - see "Before the numbers" below) bought between August 2015 and July 2016 and sold again in the year to July 2026 went for less than the previous buyer paid: 661 of the 1,673 we could match, or 39.5%. The typical (median) flat rose 4.3% in cash terms over the decade. London terraced houses, measured the same way, rose a median 27.5%, and only 6.2% sold for less. So the belief is roughly right for the typical London flat - most still sold for more, but only just. In the three other cities we tested, flats rose far more, so on this evidence it looks like a London story rather than a flats story.
Before the numbers: "London" here is the London post town (the E, EC, N, NW, SE, SW, W and WC postcode areas), not Greater London, so outer boroughs such as Croydon and Harrow are not included. Prices are cash, not adjusted for inflation. Only flats that sold in both periods are counted, and the data records prices, not reasons. Full caveats are below.
How we tested it: the same flat, sold twice
Comparing the average flat sold in 2015 with the average flat sold today mixes different homes in different places. So we compared each flat with itself. We took every standard sale of a London flat between 1 August 2015 and 31 July 2016 and looked for the same address - same postcode, building and flat - selling again between 1 August 2025 and 31 July 2026. Holding periods run from nine to eleven years.
| Homes | Matched pairs | Sold for less | Median change |
|---|---|---|---|
| London flats | 1,673 | 39.5% | +4.3% |
| London flats bought new in 2015/16 | 251 | 53.8% | -1.4% |
| London flats bought as a resale in 2015/16 | 1,422 | 37.0% | +5.4% |
| London terraced houses | 565 | 6.2% | +27.5% |
The median London flat in our matched set sold for £442,000 in 2015/16 and £470,000 in 2025/26.
It depends where in London
| Postcode area | Matched pairs | Sold for less | Median change |
|---|---|---|---|
| N | 235 | 25.1% | +10.3% |
| SE | 443 | 31.2% | +7.5% |
| E | 284 | 38.0% | +7.0% |
| NW | 137 | 43.1% | +3.4% |
| SW | 434 | 47.5% | +0.8% |
| W | 116 | 62.1% | -5.6% |
| EC | 20 | 85.0% | -13.7% |
EC rests on 20 pairs and should be read as a hint, not a finding. WC had four pairs, too few to show.
Outside London, flats did not stand still
We ran exactly the same test on flats in three other large post towns.
| Post town | Matched pairs | Sold for less | Median change |
|---|---|---|---|
| Manchester | 135 | 8.9% | +36.4% |
| Bristol | 197 | 10.2% | +27.3% |
| Birmingham | 90 | 16.7% | +21.0% |
| London | 1,673 | 39.5% | +4.3% |
These samples are much smaller than London's, so treat them as indicative. A post town is a postal area, not a council boundary: Manchester's includes parts of Salford and Trafford, and Bristol's takes in some surrounding towns.
What this does not show
- These are cash prices. Nothing is adjusted for inflation, so in real terms the London flat figures are weaker than they look.
- Only flats that sold twice are counted. Owners who could not sell, or chose not to, are not in the data, and we cannot tell which way that leans.
- Nothing about condition or size. A house can be extended or improved between sales more easily than a flat. That can flatter the house figures.
- The data does not say why. HM Land Registry records a price and a date, not the reason. We have not tried to explain the gap.
- "London" means the London post town - the E, EC, N, NW, SE, SW, W and WC postcode areas. Outer boroughs with their own post town, such as Croydon, Harrow, Kingston upon Thames and Romford, are not included.
- England and Wales data only. HM Land Registry Price Paid Data does not cover Scotland or Northern Ireland.
- Standard sales only. Repossessions, buy-to-let purchases HM Land Registry can identify by a mortgage, and sales to companies and other non-private buyers (its category B) are left out, so some investor purchases are not in these figures.
- Recent months are incomplete. Sales are added as they are registered, so these figures can change as late sales arrive. Flats resold in February to July 2026 show more selling for less than those resold in August 2025 to January 2026.
How we did it
- Source. HM Land Registry Price Paid Data as held in Domovita's copy on 28 October 2026 (sales recorded up to 28 August 2026).
- Standard sales only, both times. Both the earlier and the later sale are category A, which HM Land Registry describes as a "standard Price Paid entry" - a single residential property sold for value. Category B entries - which HM Land Registry says include repossessions, buy-to-let purchases where they can be identified by a mortgage, and transfers to non-private individuals such as companies - are left out, so investor purchases of these kinds are not in the figures (HM Land Registry guidance).
- Deletions removed. Records HM Land Registry has flagged as deleted are excluded from both sales.
- Flats are HM Land Registry's "Flats/Maisonettes" type. If the later sale recorded the home as a different type, the pair is dropped. "Bought new" uses HM Land Registry's newly-built flag on the 2015/16 sale.
- Addresses with a postcode only. Sales recorded without a postcode are left out of both periods, so an address cannot be matched to a different place with the same house name or number.
- One pair per sale. If a flat sold twice in either year, we keep one pair: its last 2015/16 purchase and its first 2025/26 sale. Where two 2015/16 sales of the same flat share a date, we keep the resale.
- Median, not average, so a handful of unusual sales cannot move it much.
- Checks. Leaving out pairs whose price fell below 0.4 times or rose above 2.5 times the earlier price gives 39.7% selling for less and a median of +4.2%. Leaving out the 311 pairs where the flat is identified only by a building number gives 43.2% selling for less and a median of +2.8%. Neither check weakens the result. Leaving out flats bought in March 2016, the busiest month (254 pairs, against 109 to 153 in each other month), gives 37.8% selling for less and a median of +5.3%; March 2016 on its own is 48.8% and +1.0%.
Using this data
The analysis on this page is published by Domovita Ltd under a Creative Commons Attribution 4.0 licence. You are free to reproduce the tables, including commercially, provided you credit Domovita and link to this page.
The underlying transaction data is not ours. Contains HM Land Registry data © Crown copyright and database right 2021. This data is licensed under the Open Government Licence v3.0. HM Land Registry does not endorse this analysis.
Cite this page
Domovita (2026) London flats ten years on: did prices stand still? Analysis of HM Land Registry Price Paid Data. https://domovita.co.uk/guides/london-flats-ten-years (accessed [date]).
Frequently asked questions
Have London flats gone up in value in the last ten years?
Only a little, in cash terms. Of 1,673 flats in the London post town (not the outer boroughs) sold in 2015/16 and again in 2025/26, the median rose 4.3%, and 39.5% sold for less than the previous price.
Is it the same for flats outside London?
Not in the cities we tested. Flats in the Manchester, Bristol and Birmingham post towns rose a median 21% to 36% on the same method, on much smaller samples, with between 9% and 17% selling for less.
Did new-build London flats do worse?
In this data, yes. Of 251 London flats bought new in 2015/16 and resold in 2025/26, 53.8% sold for less and the median change was -1.4%.
Can I see what flats in my building have sold for?
Yes. Our sold house prices section shows the most recent standard sales HM Land Registry has recorded for each street in England and Wales (category A entries; repossessions and other additional entries are not shown).
Related guides
- Street names and house prices in England and Wales
- A Seller's Guide to Listing Your Home in England
- Sold house prices by street, district and town
This page is general information based on published HM Land Registry data. It is not a valuation, an estimate of any particular property's value, or financial advice. Sold prices describe past transactions and are not a guide to what a home would achieve today.