EPC Statistics: What the Energy Performance Certificate Data Shows

Energy Performance Certificates are the basis of many energy efficiency policies in the domestic built environment, and the statistics built from them are the closest thing the UK has to a running census of how efficient its homes are¹. In the quarter January to March 2026, 460,000 EPCs were lodged on the register across England and Wales, domestic and non-domestic together, a 2% decrease compared with the same quarter in 2025¹. Of those, 416,000 were domestic certificates in England, a 1% decrease on the same quarter a year earlier, and 22,000 covered domestic dwellings in Wales, an 8% decrease¹.

The headline numbers matter because lodgement counts are used as a proxy for housing activity: a certificate is required when a home is built, sold or let, so the quarterly flow tracks the market rather than the stock. Over the 12 months to March 2026, 1,649,000 domestic EPCs were lodged in England, a 4% increase on the previous 12 months¹. The band distribution tells a different story, about the condition of the housing stock itself, and it is the part of the data that policy targets are written against.

This page explains what an EPC contains, how the 1 to 100 score and the A to G bands are produced, when a certificate is required, what the quarterly statistics show, how coverage of the housing stock varies, and what reform is coming.

What an EPC is and what the certificate contains

An EPC is a widely used measure of the energy performance of buildings, in both the residential and commercial sectors³. It provides information about the energy efficiency and emissions of a building, and it comes with a report setting out the cost effective measures that owners or businesses can take to improve the energy efficiency and energy rating, such as low energy bulbs or upgrading insulation⁴. The certificate itself carries information on potential energy costs and carbon dioxide emissions, with a coloured chart showing how energy efficient the building is or could be⁵.

The Scottish Government's guide to EPCs describes the same document in similar terms: buildings are rated on a scale from A to G, with A being the most efficient, and the certificate includes information on measures which could be made to improve the energy efficiency of the building, with an indicative cost for each improvement⁶. That improvement report is the part of the certificate most often overlooked, and it is the part that connects the rating to a household's own decisions about insulation, glazing or heating.

The certificate is not a statement of measured consumption. It is a modelled assessment of the building as a fixed set of characteristics, produced by an accredited assessor, and it is that modelled quality which makes it usable as a policy instrument and as a dataset. It is also why the statistics derived from it describe the stock rather than the behaviour of the people living in it.

The rating scale: 1 to 100, bands A to G

Both the energy efficiency rating and the environmental impact rating are scored on a scale of 1 to 100 and split into bands from A to G⁴. The score is typically expressed as a figure ranging from 1 to 100, and the bands map onto it as follows²:

The Scottish Government describes the same scale as a 1 to 100+ score presented on a distinctive graphic which shows how the score maps onto seven rating bands represented by the letters A to G⁷. The quarterly England and Wales release notes that assessments are banded from A to G, where A, or A+ for non-domestic properties, is the most efficient in terms of lighting, heating and hot water¹.

The band boundaries are fixed, which is what makes year on year comparison possible, but the score behind them is not a physical measurement. Two homes with the same band can sit at opposite ends of a range, and a home that moves from the top of band D to the bottom of band C has crossed a policy threshold without changing very much in absolute terms. That is worth holding in mind when reading any statistic expressed as a share of homes in band C or above.

How the rating is calculated: SAP and RdSAP

To determine the energy efficiency rating of a home, EPC assessors use the Standard Assessment Procedure, known as SAP⁴. The content of the domestic EPC is produced by calculations carried out in the SAP methodology⁸, and the procedure also supports the production of EPCs alongside its role in energy rating of dwellings⁹. For new dwellings in England, the approved version is the government's Standard Assessment Procedure for the Energy Rating of Dwellings, version 10.3, which also produces the target CO2 emission rate for the dwelling as designed and constructed¹⁰.

Existing dwellings are treated differently. EPCs for existing dwellings are produced using the Reduced Standard Assessment Procedure, or RdSAP, a simplified implementation of the SAP methodology based on a site survey, used when the complete data set for a SAP calculation is not available⁷. That distinction is the single most important caveat in the whole dataset. A certificate lodged for a Victorian terrace under RdSAP and a certificate lodged for a new build under SAP 10.3 are not produced by the same calculation, and comparing them directly compares two methodologies as much as two buildings.

The practical consequence is that the register holds certificates produced under successive versions of the methodology, and the statistics release is explicit that the underlying assessment regime changes over time. Where a figure is described as a median score, it is a median across whatever mix of assessment versions was lodged in the period, not a like for like physical measurement of the stock.

When an EPC is required, and who must provide it

EPCs are needed whenever a property is built, or marketed for sale or rent⁵. More precisely, they must be provided when a building is advertised for sale or for let to a new tenant, and they are also required upon completion of construction of a new building, and must be displayed by certain large non-domestic buildings¹¹. The Planning Portal states the same rule for existing buildings: an EPC is required when selling or letting an existing building¹². The Energy Performance of Buildings Directive as transposed in the UK requires certificates when any building is sold, rented out or constructed, and sometimes after refurbishment work¹³.

It is a legal requirement for all new build domestic dwellings in England and Wales to have an EPC when completed¹⁴. In Scotland, the Energy Performance of Buildings (Scotland) Regulations 2008 place obligations on building owners to provide an energy performance certificate when buildings are to be sold or rented out, and a requirement to issue EPC data to a register has also been introduced¹⁵. An EPC is required as part of the Home Report for buildings for sale in Scotland, and an EPC is required for every individual dwelling that is rented out¹⁶.

Responsibility sits with the owner. It is the owner's or landlord's responsibility to make sure they have an EPC, to make it available to potential buyers or tenants, and to give a copy to the new owner when selling⁵. Private rented sector properties in England and Wales must meet a minimum EPC rating of E or above, and the government publishes the exemptions from that standard¹⁸.

How many EPCs are lodged each quarter

Lodgement counts are the flow measure in the statistics, and they move with the housing market rather than with the stock. The most recent release covers January to March 2026¹:

The 12 month figures smooth the quarterly noise. In the 12 months to March 2026, 1,649,000 domestic EPCs were lodged in England, a 4% increase on the previous 12 months, and 204,000 EPCs were lodged for new dwellings, a 2% increase on the previous year¹. Non-domestic lodgements ran the other way: 87,000 in the 12 months to March 2026, an 8% decrease on the previous 12 months¹.

For comparison, the April to June 2025 quarter saw 495,000 domestic EPCs lodged in England, a 21% increase on the same quarter a year earlier, and 544,000 EPCs lodged across England and Wales including non-domestic certificates¹⁹. Non-domestic lodgements in that quarter were 23,000 in England, a 9% decrease compared with the same quarter in 2024, and 93,000 over the 12 months to June 2025, a 14% decrease¹⁹. The gap between a 21% rise in one quarter and a 1% fall a few quarters later is the clearest illustration that this series tracks transaction and construction activity, not a steady underlying trend.

What the ratings show: how efficient the housing stock is

The band distribution is where the statistics become a statement about the country rather than about the market. New dwellings are consistently the most efficient part of the stock: in the quarter January to March 2026, 87% of new dwellings in England were given an A or B rating, an increase from 86% in the same quarter a year earlier¹. The median score for new dwellings in England rose from 82 in the five year period to March 2013 to 84 in the five year period to March 2025².

Older and rural stock sits lower. Flats and maisonettes were the most energy efficient property type in England and Wales, with a median score of 74, while existing larger rural properties scored 67, similar to urban or larger rural areas, both scoring 69². The gap between a median of 84 for new build and 67 for larger rural existing homes is the stock's central fact: efficiency is concentrated in what has been built recently.

The longer trend is one of improvement. In 2010, just 14% of UK homes had an Energy Performance Certificate of C or above; by the time of that statement it was 46% and rising²⁰. In Scotland, homes in the social rented sector are now some of the most energy efficient, with 85% already achieving EPC D or above²¹. An earlier Scottish assessment found social housing generally more energy efficient, with 56% of homes having attained EPC C²².

Ratings by tenure, dwelling type and age

Tenure is one of the sharpest dividing lines in the data. Around 2% of dwellings in the social sector are within EPC bands E, F or G, while 12% of owner-occupied dwellings and 14% of the private rented sector are within those bands²³. The private rented sector therefore has the largest share of the least efficient homes, which is the reason minimum standards have been aimed at it.

Age and location reinforce the pattern. New dwellings in England moved from a median score of 82 in the five year period to March 2013 to 84 in the five year period to March 2025, while existing larger rural properties scored 67 and urban or larger rural areas scored 69². Coverage itself is uneven by age: a much smaller proportion of dwellings built before 1930 in Wales were covered by an EPC, at 58%².

Northern Ireland publishes its own breakdown. The Northern Ireland Statistics and Research Agency table CT0059 presents Energy Performance Certificate band by household tenure by resident characteristics, using Census 2021 estimates that classify usual residents in Northern Ireland living in households by EPC band, by household tenure and by resident characteristics²⁴. That is a different construction from the England and Wales lodgement series: it is a census based classification of households rather than a count of certificates lodged, and the two should not be read as the same measure.

Coverage: how much of the housing stock has an EPC

Coverage is the question that determines how far the statistics can be treated as a description of the whole stock. Just over 71% of all residential dwellings in England, and 67% in Wales, had at least one EPC registration since records began in 2007, as of March 2025². The quarterly release puts the same measure at around 71% of the housing stock in England and around 67% in Wales, based on Census 2021 analysis¹.

An earlier release gave lower figures, describing EPCs as covering around 60% of the housing stock in England and around 55% in Wales, also based on Census 2021 analysis by the ONS¹⁹. The two sets of figures are not reconciled in the documents, and the difference is large enough to matter when coverage is used to judge how representative the band distribution is. Both are official statistics; the later release gives the higher figures.

Coverage is also uneven within each nation. A much smaller proportion of dwellings built before 1930 in Wales were covered by an EPC, at 58%². Since older homes are also the least efficient on average, the properties least likely to appear in the data are the ones the data would most usefully describe. That is a structural limitation of a register built from transaction and construction events rather than from a survey of the stock.

For context on the denominator, the dwelling stock estimates for England are accredited official statistics, published for 31 March 2025²⁵. The gap between the number of dwellings in England and the number with a lodged certificate is the part of the housing stock for which no EPC based assessment exists at all.

What the data is used for, and who can access it

The EPC is publicly available, and some of the underlying data may be shared with others for the monitoring and compliance of the EPB regulations, with government also able to use some data for research or statistical purposes⁵. The register itself is kept by the Secretary of State for the Department of Levelling Up, Housing and Communities, and the unauthorised disclosure of data is a criminal offence²⁶.

User research into the England and Wales statistics found the data is generally split between several needs: communication methods for journalist or social media posts, comparing areas of housing to help understanding of possible improvements, gathering data on rating and floor area to create price per metre data, and enabling up to date calculation of the carbon score²⁷. The same research found the most useful data is the current EPC, its potential and the floor area, and noted a dashboard to provide more accessibility to the data underlying the live tables²⁷. Seven users were interviewed remotely for that study²⁷.

The statistics themselves have been upgraded in status. The figures in the release were changed from "Official Statistics in development" to "Official Statistics" from October 2024, on the basis of improvements to the data¹. The next release, covering April to June 2026 for England and Wales, is expected on 30 July 2026¹.

Reform: new ratings, shorter validity and the Home Energy Model

The assessment regime is being rebuilt. The Scottish Government has set out plans to reform EPCs by introducing new ratings, redesigning the certificates, and making improvements to the operational infrastructure¹¹. The UK Government expects the Home Energy Model to be introduced as the basis for calculating EPC ratings in the second half of 2026, establishing a new technical infrastructure to replace the current Standard Assessment Procedure with the new UK Home Energy Model, together with a new EPC Register²⁸.

In Scotland, EPCs are to be generated using the UK Government's new Home Energy Model, which will replace the current Standard Assessment Procedure methodology in autumn 2026²⁹. Validity is also in scope: the Scottish Government has stated that it is reducing the validity period of EPCs from ten to five years²⁹. The current rule remains ten years, and a certificate can be used multiple times during that period⁵.

Policy targets are written against the bands. Upgrading fuel poor homes in England to EPC band C "where reasonably practicable" is one stated aim³⁰. The government has also announced £15 billion to improve energy efficiency and cut bills²⁰. Reducing the price of electricity relative to gas would make heat pumps more competitive, which bears on the running cost case for low carbon heating rather than on the certificate itself³¹.