Energy ratings,
through time.
Compare the first and latest recorded grades of properties assessed more than once in England and Wales.
England and Wales
Recorded assessments from 21 Sept 2008 to 31 Oct 2026. Dates differ between properties.
First and latest grades
Latest recorded gradeCell intensity uses a logarithmic scale for this cohort. An asterisk marks a withheld count below 10.
Selected transition
Higher recorded grade
1,265,94716.53% of this cohort
This compares the first and latest usable assessments for each property. It does not establish what caused the change.
Read the change
The same properties.
Different dates.
This cohort includes properties with usable domestic certificates on at least two different dates. Each property contributes its first and latest recorded grade, so the time between assessments differs from one home to another. This is a sample of repeatedly assessed properties, rather than the whole housing stock.
A higher recorded grade does not establish that renovation caused the change. Assessment methods, recorded details and boundaries can change. Conflicting grades recorded for the same property on the same date are excluded.
Areas follow the latest usable certificate. Properties without a usable local area appear only in the national cohort. Certificate dates use lodgement first, then registration or inspection when the earlier date fields are absent. Groups smaller than 10 are withheld; a displayed zero means no properties in this cohort have that grade combination.
Source: Energy Performance of Buildings Data, updated 1 November 2026. Analysis cutoff 8 November 2026. Contains public sector information licensed under the Open Government Licence v3.0. Source licence.
Read the methodologyLook inside the assessment
Heating, walls
and roofs.
Compare the recorded feature ratings at the same first and latest assessment dates. These ratings do not prove that insulation was installed or that equipment is working.
England and Wales
Paired assessments from 21 Sept 2008 to 31 Oct 2026. Dates vary between properties.
Higher, same and lower shares describe the paired cohort with a known rating at both endpoints.
Endpoint coverage
7,656,746 eligible repeat propertiesUnknown includes missing, unrecognised, N/A and conflicting endpoint values. These coverage counts overlap.
First and latest ratings
Latest recorded ratingOnly pairs with two known ratings appear in the matrix. Cell intensity uses a logarithmic scale. An asterisk marks a withheld count below 10.
Properties with these ratings
Higher recorded rating
550,6317.24% of the paired cohort
A recorded rating change does not establish what work took place or what caused it.
About these assessments
The official scale runs from very poor to very good. These are modelled EPC feature assessments and can be based on assumptions. Assessment methods can change between certificates.
Each feature uses the same earliest and latest certificate dates as the grade comparison. An unknown endpoint is not replaced with a known rating from another date. Properties with conflicting grades on the same certificate date are excluded from the repeat cohort.
Areas follow the latest usable certificate. Missing local areas remain in the national cohort only. The comparisons cover England and Wales and do not represent all homes.
Source: Energy Performance of Buildings Data, updated 1 Nov 2026. Analysis cutoff 8 Nov 2026. Contains public sector information licensed under the Open Government Licence v3.0. Source licence. Official feature definitions. Insulation interpretation.