Twenty-nine of England and Wales's busiest city-centre postcodes are overwhelmingly flats — several are 95% or more apartments, with houses almost absent. But "mostly flats" turns out to hide two very different housing markets. In some centres, that flat stock is being built right now. In others, it is decades old, and almost nothing new has been added in years. HM Land Registry's Price Paid Data lets the two be told apart, because every sale carries a new-build flag alongside the property type.
Pooling the two most recently complete years of sales data (2023 and 2024, chosen to avoid a registration-lag distortion explained below), 7,672 flats sold across these 29 postcodes — and only 1,733 of them, 22.6%, were new-build. The rest, more than three in four, were resale: flats changing hands for the second, third or tenth time, not freshly completed stock reaching the market for the first time.
Six near-identical flat shares, six different stories
The clearest way to see this is to isolate the centres where flats are close to the entire market. Six of the 29 postcodes had a flat share above 94% in 2023–24: Manchester (M1, 99.2% flats), Newcastle (NE1, 96.7%), Liverpool (L1, 96.6%), Leeds (LS1, 96.4%), Sheffield (S1, 96.0%) and Bristol (BS1, 94.8%). On flat share alone, these six look like the same kind of place — a city core with almost no houses left.
Their new-build shares say otherwise. Liverpool's L1 flats were 58.4% new-build over the two years; Manchester's were 49.4%; Sheffield's were 46.8%. Newcastle's NE1 flats, despite being just as flat-dominated as Manchester's, were only 1.7% new-build — and Bristol's BS1 sat at 13.3%. A buyer comparing Manchester and Newcastle city centres on "almost entirely flats" alone would miss that one market is in the middle of a sustained construction programme and the other is trading much the same apartment stock it had a decade ago.
The full ranking
All 29 centres, ranked by the share of their flat sales that were new-build in 2023–24 (category-A Land Registry sales, property_type = 'F', new_build = 1):
| City (postcode) | Flat share, 2023–24 | Flats sold | New-build share of flats |
|---|---|---|---|
| Derby (DE1) | 59.3% | 267 | 62.5% |
| Liverpool (L1) | 96.6% | 481 | 58.4% |
| Manchester (M1) | 99.2% | 622 | 49.4% |
| Sheffield (S1) | 96.0% | 218 | 46.8% |
| Hull (HU1) | 85.8% | 121 | 45.5% |
| Birmingham (B1) | 94.2% | 372 | 45.4% |
| Preston (PR1) | 17.7% | 274 | 40.5% |
| Cambridge (CB1) | 26.1% | 279 | 21.5% |
| Reading (RG1) | 52.6% | 619 | 18.6% |
| Leeds (LS1) | 96.4% | 81 | 18.5% |
| York (YO1) | 71.8% | 130 | 15.4% |
| Bradford (BD1) | 93.9% | 214 | 14.0% |
| Bristol (BS1) | 94.8% | 346 | 13.3% |
| Norwich (NR1) | 49.9% | 449 | 12.5% |
| Bath (BA1) | 37.9% | 471 | 12.1% |
| Stoke-on-Trent (ST1) | 9.2% | 66 | 10.6% |
| Exeter (EX4) | 21.9% | 267 | 10.1% |
| Plymouth (PL1) | 69.1% | 511 | 8.6% |
| Cardiff (CF10) | 88.8% | 426 | 7.7% |
| Southampton (SO14) | 71.0% | 348 | 6.9% |
| Nottingham (NG1) | 88.6% | 148 | 2.0% |
| Newcastle (NE1) | 96.7% | 174 | 1.7% |
| Swansea (SA1) | 24.4% | 253 | 0.4% |
| Coventry (CV1) | 36.2% | 92 | 0.0% |
| Leicester (LE1) | 91.3% | 116 | 0.0% |
| Middlesbrough (TS1) | 8.1% | 18 | 0.0% |
| Oxford (OX1) | 36.8% | 119 | 0.0% |
| Portsmouth (PO1) | 36.3% | 147 | 0.0% |
| Wolverhampton (WV1) | 24.4% | 43 | 0.0% |
Six of the 29 — Coventry, Leicester, Middlesbrough, Oxford, Portsmouth and Wolverhampton — recorded not a single new-build flat sale in either year. Their flat markets, whatever size they are locally, were entirely resale across this window.
Where the building boom is real
Derby's DE1 postcode is the single biggest outlier: only 59.3% flats (it still has a meaningful terraced-house market), yet 62.5% of those flats sold in 2023–24 were new-build — the Castleward regeneration scheme by the river is still actively completing and selling. Liverpool, Manchester and Sheffield each combine a near-total flat market with new-build shares comfortably above 45%, consistent with the apartment-block and warehouse-conversion-plus-new-build pipelines that have run through all three city centres this decade. Hull and Birmingham sit just behind them. In these six places, a large and recent construction pipeline is doing real work in both the flat share and the flat mix.
Where the flat market is simply old
Newcastle is the sharpest contrast in the data set. NE1 is 96.7% flats — as flat-dominated as Manchester — but only 1.7% of those flat sales were new-build across two full years. The postcode's apartment stock is overwhelmingly older conversions and purpose-built blocks changing hands again and again, not a recent development wave. Bristol's BS1 (13.3% new-build), Nottingham's NG1 (2.0%) and Leicester's LE1 (0.0%, despite 91.3% flats) tell variants of the same story: a flat-dominated core is not, on its own, evidence of recent building. It can just as easily mean a market that finished building some years ago and has been reselling the same stock since.
A weak relationship, citywide
Across all 29 centres, the correlation between a postcode's flat share and its new-build share of flats is weak (Pearson r = 0.35 over the two-year pool) — positive, but nowhere near strong enough to treat flat-dominance as a proxy for "new-build city centre." Knowing a centre is nearly all flats tells you almost nothing about how much of that stock is new; the new-build flag has to be checked separately. This follows a companion finding on these pages that flat share itself barely tracks city-centre prices (Pearson r = −0.19) — on this data set, a city centre's flat-vs-house mix is a surprisingly poor stand-in for either of the two things it is often assumed to signal.
Why 2023 and 2024, not the latest year available
This analysis deliberately stops at 2024 rather than extending into 2025. Land Registry sales take months to register after completion, and that lag falls hardest on new-build flats specifically — large apartment schemes often complete and register in stages well after the headline sale date, so the most recent year in any rolling data set understates new-build activity more than it understates resale activity. Pooling the two most recently complete years smooths both that registration lag and the lumpiness of individual development completions, which can otherwise make a single year look more (or less) new-build-heavy than the underlying market really is.
None of this is a forecast. It describes what was recorded as sold across two specific years, drawn directly from HM Land Registry's Price Paid Data as held in Homecost's database — not an estimate, and not a listing site's snapshot of what is currently for sale.
Homecost's postcode tool shows the sale history behind any of these figures directly — Manor Chare Apartments in Newcastle's NE1 is a good illustration of the city's older, resale-dominated flat stock, with recorded sales stretching from 2004 through to early 2026 and no new-build flag among them. For the companion finding on which centres have almost no houses left at all, see the UK city centres with almost no houses; for how flat share has shifted since 2020 across the same 29 postcodes, see city-centre flat shares since 2020; and for how flat share relates to price across the same set, see flat share versus city-centre prices. More market analysis is on homecost.uk.