Which UK Regions Have the Worst EPC Ratings? A Data Look at England’s Least Energy-Efficient Homes

Every quarter, the government publishes a fresh batch of data on the energy efficiency of homes across England and Wales. Buried in that data - spread across the English Housing Survey, the Office for National Statistics’ analysis of the Energy Performance of Buildings Register, and the Department for Energy Security and Net Zero’s fuel poverty statistics - is a clear regional pattern: some parts of England are lagging years behind others when it comes to insulation, heating efficiency and overall EPC performance. This piece pulls together what the official data actually says, names the regions and local authorities it points to, and looks at why the gap exists. Nothing below is estimated or inferred beyond what the source publications state.

One scoping note before the numbers: this piece focuses on England. Scotland, Wales and Northern Ireland each run their own EPC register and publish their own statistics, using methodology that isn’t directly comparable to England’s - the Scottish EPC register, for example, is administered separately from the England and Wales register, and the House of Commons Library’s own briefing on energy efficiency notes that Northern Ireland and Scotland’s ratings aren’t calculated on an identical basis to England’s. Wales shares a joint EPC register with England and appears in the same quarterly statistical releases, but its housing stock and its results are different enough - Wales’ median EPC score sits at 68, in band D, according to the ONS’s 2025 analysis - that it’s treated separately here rather than folded into “UK” or “England” figures.

What the government’s own housing data actually shows

The starting point is the English Housing Survey, the government’s official annual survey of the condition of the housing stock in England, run by the Ministry of Housing, Communities and Local Government (MHCLG). Its energy efficiency findings are usually treated as the gold-standard measure of the whole stock, because - unlike the Energy Performance Certificate (EPC) register, which only captures homes that have been sold, let, or newly built since EPCs became mandatory - it samples the entire English housing stock, including homes that have never had a certificate at all.

A quick note on units, since two different scales turn up in the data below and it’s easy to confuse them. EPC bands (A to G) are the familiar letter grades, but they’re derived from an underlying numerical SAP score running roughly from 1 to 100-plus, where band C covers the range from 69 to 80 and band D covers 55 to 68. When the ONS or the government report a “median score” of, say, 68 or 70 for a region, that’s this underlying numerical score, not a percentage - a median of 70 means the typical home in that region sits right at the top of band C, while a median of 68 sits at the very top of band D. It’s a small distinction, but it explains why a two-point gap in median score between the best and worst English regions can still coincide with a much wider gap in the share of homes rated D or below.

The headline trend is genuinely positive. According to the English Housing Survey, the proportion of English homes rated band A to C rose from 26% in 2014 to 48% in 2022, and reached 56% by 2024. In other words, a majority of English homes are now rated C or above for the first time on record. If you’re not familiar with what these letter bands actually represent in terms of insulation, heating costs and building fabric, it’s worth reading our explanation of what EPC ratings actually mean before going further, since the rest of this piece assumes you know roughly what separates a band C home from a band E or F one.

That national improvement, though, hides a wide regional spread. The most detailed regional breakdown currently published by the English Housing Survey covers 2022 data (published July 2024) - the 2024 headline figures above are more recent, but the detailed regional tables for that later survey year have not yet been released in full. Working from the most recent available regional breakdown, the pattern is stark and consistent with what the separate ONS analysis of the EPC register finds too.

Which English regions have the worst EPC ratings

Based on the English Housing Survey’s 2022 regional data, the share of homes rated band A to C varies by nearly 20 percentage points across England’s nine regions:

The South West stands out clearly as the weakest-performing region in this dataset, with 15% of its housing stock rated E, F or G - the highest proportion of poorly-rated homes of any English region, and notably worse than the roughly 4% to 12% seen elsewhere. London, by contrast, comes out on top, helped by its higher proportion of flats and more recently-built stock (more on why property type and age matter so much below).

One result in that list is genuinely counter-intuitive: the North East ranks as the second-best region in England for EPC ratings, ahead of the South East, both Midlands regions, and the whole of the North West. This is a useful reminder that “worst EPC ratings” doesn’t map neatly onto simple assumptions about wealth or the North/South divide - the South West, a region with a reputation for affluent market towns and seaside prosperity, actually has the worst-rated housing stock in the country by this measure.

A separate, independently-run piece of analysis backs up the broad shape of this picture. The Office for National Statistics’ review of the Energy Performance of Buildings Register - which looks at actual lodged EPC certificates rather than a survey sample - found that homes in London, the East of England and the South East shared the highest median EPC score of all English regions, at 70 points (the upper end of band C), based on records lodged between April 2015 and March 2025. Yorkshire and the Humber had the lowest median score of the English regions at 68 (band D) in the same analysis, though the ONS was careful to note this is a narrow two-point spread across the country as a whole - most of the regional gap in England shows up in the tails (the best and worst local authorities) rather than in the regional averages themselves.

Beneath the regional averages: the local authorities dragging the numbers down

Regional figures smooth over a lot of local variation, and this is where the government’s own data gets more interesting. The ONS’s analysis of the EPC register broke performance down to individual local authorities and even to MSOAs (small statistical areas covering a few thousand households each), and found genuinely wide gaps within regions, not just between them.

At the top end, Tower Hamlets in London had the highest proportion of any local authority of homes with a valid EPC scoring 69 or above (band C or better) - 82% of its certified stock. Knowsley in the North West was the next best-performing local authority, at 74%. Both are dominated by newer or purpose-built flatted stock, which tends to score well on EPCs regardless of region.

At the other end, the same ONS analysis identified six English local authorities where not a single MSOA achieved an average EPC score of band C or above across the whole area - meaning every single small statistical sub-area within these local authorities averaged band D or worse:

Three of those six - Hyndburn, Pendle and Rossendale - sit within a few miles of each other in East Lancashire, which points to a genuine local cluster of older, harder-to-treat housing stock rather than a coincidence of boundary lines.

Separately, several local news outlets and trade publications have run their own analysis of the same underlying open EPC register (the government’s Energy Performance of Buildings Data, published via opendatacommunities.org) at local authority level, and their figures are consistent with the pattern the ONS describes. Local reporting using this open register has put the proportion of homes rated F or G - the two worst bands, both below the legal minimum standard for most rented properties - at around 17% in Eden, Cumbria, over 15% in West Devon, and around 14% across Cornwall as a whole; a Yorkshire local paper reported a comparable 15% F-or-G figure for Ryedale in North Yorkshire, using the same public dataset. These are third-party calculations from the open government register rather than headline gov.uk statistics in their own right, but they’re drawing on the same official source and the same methodology the ONS itself uses, and the local authorities they flag - largely rural, with older and more dispersed housing stock - fit the pattern the national data shows.

Why some regions score so much worse than others

The government’s data is fairly clear on the main drivers, and none of them are really about geography for its own sake - they’re about what kind of housing stock happens to sit where.

Age of the building is the single biggest factor. The English Housing Survey’s 2022 data shows only 21% of pre-1919 homes are rated band A to C, compared with 83% of homes built after 1990. Nearly a quarter (23%) of pre-1919 homes are rated E, F or G. The ONS’s separate analysis of the EPC register found the same effect from a different angle: properties built after 2012 had a median EPC score of 84, compared with a much lower median for anything built before 2012, and only 61% of pre-1930s properties in England even have a valid EPC on record at all (against 94% for post-2012 builds) - meaning the true picture for the oldest housing stock, much of which has never needed a certificate, may be even less complete than the register suggests.

Property type compounds the age effect. Flats are consistently the best-performing property type nationally - the English Housing Survey found 83% of purpose-built high-rise flats and 72% of low-rise flats rated A to C, against 47% for detached houses, 45% for semi-detached, and just 32% for terraced houses. Converted flats (typically older houses split into separate units) performed worst of all house-type categories, with only 18% rated A to C. Regions with a higher share of flats in their housing mix - London above all - get a systematic uplift in their averages that has nothing to do with better retrofitting and everything to do with what was already there.

Rural areas underperform urban ones. The ONS’s EPC register analysis found smaller rural properties had a median score of just 63, against 69 for both urban areas and larger rural settlements - and for existing (non-new-build) smaller rural properties specifically, the median dropped to 59. This matters for how the regional rankings should be read: the South West and Yorkshire and the Humber both contain large stretches of dispersed rural housing, off the gas grid in many cases, which drags their averages down in a way that a more urbanised region with the same building age profile wouldn’t experience.

Tenure plays a role too. Housing association homes performed best of any tenure in the English Housing Survey data, at 72% rated A to C - better than both private rented (45%) and owner-occupied (43%) homes. This is generally attributed to the sustained programme of retrofit investment social landlords have been required to make, compared with the much more piecemeal, homeowner-led pace of improvement in the private and owner-occupied stock - which happens to be the vast majority of homes in lower-density, rural and semi-rural regions.

None of these factors are evenly spread across England, which is exactly why the regional table above looks the way it does. The South West and the East Lancashire cluster of local authorities named earlier both combine several of the worst factors at once - older stock, a lower proportion of flats, more rural and off-gas-grid housing, and a higher share of owner-occupied and private-rented homes relative to housing association stock. London and the North East, by contrast, each benefit from one or more of these factors working in their favour, whether that’s London’s stock of flats and new-build or the North East’s particular mix of housing types and tenures. The regional label is really just a proxy for the underlying mix of building age, type, tenure and rurality that the English Housing Survey measures directly.

The fuel poverty connection: bad EPC ratings aren’t just an efficiency problem

The government doesn’t just track EPC ratings for their own sake - it uses them directly to define fuel poverty. Under the Low Income Low Energy Efficiency (LILEE) metric used in England, a household is officially classed as fuel poor if it lives in a home rated band D, E, F or G and its income (after housing and energy costs) falls below the poverty line. That means the regions with the worst EPC ratings and the regions with the worst fuel poverty rates are, by definition, connected.

The government’s most recent sub-regional fuel poverty statistics (2024 data) show fuel poverty rates ranging widely across England, with a national average of 9.9%:

At local authority level, the same report put the five worst-affected areas as the Isles of Scilly (17.3%), Stoke-on-Trent (14.8%), Blackpool (13.9%), Birmingham (13.8%) and Sandwell (13.5%) - while the five least-affected were Wokingham (5.6%), Bracknell Forest (5.8%), Surrey Heath (5.8%), Hart (6.0%) and Eastleigh (6.0%), all in the South East’s commuter belt. The government’s own analysis attributes the West Midlands’ high rate mainly to lower median household income rather than a dramatically worse housing stock than other regions, which is a useful corrective: a poor EPC rating becomes a fuel poverty problem specifically where it collides with low income, not just where the building fabric happens to be old.

It’s worth noting the Isles of Scilly turns up at the bottom of both the EPC league table (as one of the six local authorities with no MSOA reaching band C) and the fuel poverty league table (the single worst local authority in England on that measure) - a small, isolated, largely older housing stock, off the mains gas grid, is about as unfavourable a combination as the English data shows anywhere.

What this means if you live in one of these regions

None of this data is really about blame - it’s about where the retrofit need is greatest, and it lines up with common sense once you see the pattern: older housing stock, rural and off-grid locations, a high proportion of solid-wall terraces or converted properties, and lower-income areas. If you live in the South West, Yorkshire and the Humber, the West Midlands, or in a rural or coastal local authority that isn’t already on the government’s radar for a fuel poverty scheme, there’s a reasonable chance your home falls into exactly the category this data describes - and a reasonable chance that means real, ongoing costs on your heating bills that a newer or better-insulated home simply doesn’t carry.

The flip side of a bad regional average is that the improvements that move a home from band E or D up to band C tend to be well understood and well documented for exactly this kind of older, solid-wall or off-grid housing stock - loft and cavity or solid-wall insulation, glazing upgrades, and a heat pump or more efficient heating system where mains gas isn’t an option. Whichever region you’re in, the first useful step is finding out exactly where your own property currently sits and what would move it up a band.

In practice, the specific fix depends heavily on which of the factors above applies to your own home, which is part of why blanket advice doesn’t travel well between regions:

Check your property’s EPC rating and get matched with a local specialist →

Blog