The UK's Most Comprehensive
EPC Dataset
EPCGuide analysed every domestic energy performance certificate ever issued in England and Wales to reveal the true scale of the UK's energy efficiency crisis. The results are published here for free: interactive maps, area rankings, and downloadable data.
EPCGuide is an independent research platform focused on making UK energy performance data accessible. We are not affiliated with any government body, energy company, or political organisation.
29.2M
certificates analysed
346
local authorities
55.3%
of homes below EPC C
The Numbers
Scale of the UK's Energy Efficiency Crisis
Derived from the complete EPC register. Every number below is calculated from the same underlying dataset of 29.2M certificates.
29.2M
EPC certificates analysed
Every domestic certificate lodged in England and Wales
55.3%
of homes are below EPC C
Fail the proposed 2030 minimum standard for rental properties
16.2M
homes need upgrades by 2030
Rated D, E, F, or G on the energy performance register
£111.7B
estimated total upgrade cost
Based on industry average retrofit costs per rating band
346
local authorities covered
Every LA in England and Wales with published EPC data
11,914
upgrades needed every day
To reach full compliance by January 2030
Worst performing area
Isles of Scilly
85% of homes below EPC C
Best performing area
Tower Hamlets
24.9% of homes below EPC C
Our research
Original Analysis. Free to Cite.
Each analysis examines a different dimension of the UK energy efficiency crisis. All findings can be cited, all data can be downloaded.
Interactive UK EPC Map 2026
The most detailed choropleth map of EPC compliance across England and Wales. Every local authority colour-coded by non-compliance rate, with sortable tables, property-type breakdowns, and CSV export.
- Hover any area for instant stats
- Sort by compliance rate, upgrade cost, or records
- Drill into property type and age band data
UK's Most Expensive Areas to Heat
Original analysis of actual heating cost data from EPC certificates reveals a stark north-south divide in energy bills. See the off-gas penalty, the age tax on Victorian properties, and total savings available.
- Ranked by average annual heating cost
- Fuel type breakdown: oil vs gas vs electric
- Victorian "age tax" quantified
EPC Rankings by Parliamentary Constituency
How does your MP's constituency rank for energy efficiency? This analysis maps EPC compliance to the 650 parliamentary constituencies in England and Wales, giving a political lens on the housing energy crisis.
- Search by constituency or MP name
- Party-by-party compliance breakdown
- Best and worst performing seats ranked
Rental vs Owner-Occupied: The EPC Gap
A counterintuitive finding: owner-occupied homes have a higher below-C rate than private rentals. But for flats, renters are measurably worse off. This analysis unpacks the split-incentive problem that leaves tenants paying for landlords' inaction.
- Surprising truth: owners underperform renters overall
- Flat-specific gap: renters 24% more likely to fail
- Social housing as the benchmark others ignore
State of EPC Compliance: City by City
City-by-city breakdown of EPC compliance rates drawing on the full 29.2 million record dataset. Includes national leaderboards by property type, upgrade cost estimates, and the scale of work needed before the 2030 deadline.
- 29.2M certificate dataset
- National leaderboards by property type
- Cost and timeline to 2030 compliance
How We Built This Dataset
Full transparency on our data sources, collection methods, cost modelling assumptions, and limitations. Journalists and researchers can cite these findings with confidence.
Data Source
All data is sourced from the official EPC Register operated by the Ministry of Housing, Communities and Local Government (MHCLG), accessed via the Open Data Communities API at epc.opendatacommunities.org.
The EPC Register is the statutory database of energy performance certificates required under the Energy Performance of Buildings (England and Wales) Regulations 2012. It contains every domestic EPC issued since 2008.
Data accessed and processed: 30 March 2026. The register is updated daily; our dataset reflects a complete pull as of that date.
What We Analysed
Our full dataset contains 29,214,082 domestic EPC certificates across 346 local authorities in England and Wales.
Where multiple EPCs exist for the same property address, we use the most recent certificate. This means our figures represent the current best-known energy performance of each property, not a count of all certificates ever issued.
Certificates with missing or invalid rating data were excluded. Properties with commercial, industrial, or non-domestic use classifications were excluded; the dataset covers residential dwellings only.
29.2M
total records
346
local authorities
2008
data from year
How We Define “Below EPC C”
A property is classified as “below EPC C” if its current EPC rating is D, E, F, or G (an energy efficiency score below 69 on the Standard Assessment Procedure scale).
This threshold reflects the government's proposed minimum energy efficiency standard for rental properties by 2030, as set out in the Warm Homes Plan consultation and the proposed update to Minimum Energy Efficiency Standards (MEES).
Properties rated C (69-80) or above are considered compliant with the proposed 2030 standard. Properties rated A (92+) or B (81-91) are the best performers.
Upgrade Cost Estimates
Total upgrade cost estimates are calculated using industry-average retrofit cost figures. We apply a per-property cost based on the gap between the current EPC rating and EPC C:
| Current rating | Target | Assumed avg cost |
|---|---|---|
| D (55-68) | EPC C | £5,500 |
| E (39-54) | EPC C | £8,500 |
| F (21-38) | EPC C | £12,000 |
| G (1-20) | EPC C | £18,000 |
Sources: DESNZ retrofit cost analysis, RICS cost guidance, Energy Saving Trust surveys 2024-25. Costs are indicative averages; actual costs vary significantly by property type, construction, and region.
Statistical Notes and Limitations
EPC coverage: EPC certificates are required at point of sale or new tenancy. Properties that have never been sold or rented since 2008 may not have a certificate on the register. The register therefore underrepresents older owner-occupied housing stock that has not changed hands recently.
Currency of assessments: EPC certificates are valid for 10 years. Some certificates in our dataset were issued up to a decade ago and may not reflect improvements made since assessment. Properties that have been retrofitted but not reassessed will appear worse than they are in reality.
SAP methodology: EPC ratings are calculated using the Standard Assessment Procedure (SAP), a modelled estimate of energy performance based on assessed characteristics. They do not measure actual energy consumption. The introduction of RdSAP 10 in 2025 changed some calculation assumptions.
Upgrade cost estimates: Total upgrade cost figures are indicative only. They assume all non-compliant properties require a single retrofit package based on average costs. Many properties will be cheaper or more expensive to upgrade depending on construction type and current features.
Cite Us. Download the Data. Tell the Story.
All research on EPCGuide is original, independently conducted, and free to cite. Data is sourced from the official government EPC register. Every analysis includes full methodology notes, CSV downloads, and a clear data provenance trail.
Suggested citation format: “EPCGuide EPC Research Hub (30 March 2026). Analysis of 29,214,082 EPC certificates. Available at: epcguide.co.uk/research”
Download CSV Data
All research pages offer raw data downloads. Local authority tables, heating cost rankings, and constituency data all available.
Start with the EPC mapOfficial Source
All data originates from the MHCLG EPC Register, the statutory source of energy performance data in England and Wales.
View the registerFull Methodology
Every calculation, assumption, and limitation is documented above. Nothing is hidden.
Read methodologyComplete Coverage
All 346 local authorities in England and Wales are covered. This is not a sample.
Explore coverage