Key takeaways
- Price Paid Data covers sales in England and Wales lodged for registration, and the latest two months are always incomplete[1].
- Asking prices are not evidence: most agents report homes selling below asking[2].
- Pick three to six comparables on the same street type, tenure and property type, then adjust for time, size and condition in writing.
- The data does not separate bungalows or record condition, so you must add what you saw on the visit.
- Keep the comparables on file; redress scheme codes expect your figure to be backed by evidence[3].
What the price paid data contains
HM Land Registry's Price Paid Data tracks residential sales in England and Wales that are sold for value and lodged for registration[1]. Each record gives the price, the date of transfer, the postcode and address, a property type code, whether the home was new build and whether it was freehold or leasehold. You can download one file covering everything since 1995, a monthly file of what was received that month, or yearly files arranged by the date of the deed[1].
Two details matter most for valuers. Property type is recorded only as detached, semi-detached, terraced, flat or maisonette, or other, so bungalows are not identified separately and end-of-terrace houses sit inside terraced[1]. And because registration lags the sale, the data for the two most recent months is incomplete[1]. A quiet-looking last quarter on a street often means the sales have not been lodged yet, not that nothing sold.
What it leaves out
The dataset excludes several kinds of transfer that would distort a valuation if they were included[1]. Knowing the list stops you hunting for a sale you heard about at the school gate and cannot find.
- Sales not lodged with HM Land Registry, and transfers that were not for value, such as gifts.
- Right to buy sales at a discount, and transfers of a share in a property, for example between a divorcing couple.
- Transfers under a compulsory purchase order or a court order.
- Leases of seven years or less, which are not recorded at all[1].
Repossessions and buy-to-let purchases with an identifiable mortgage appear as Category B entries, separate from standard Category A sales[1]. For a family home appraisal, filter to Category A first and only look at Category B if you need to explain an unusually low figure on the street.
Why asking prices are not comparables
Portal asking prices tell you what sellers hope for. In August 2026, 86% of Propertymark member agents reported homes selling for less than the asking price and only 9% at asking[2]. The average asking price of a newly listed home in Great Britain was £371,015 in October 2026, almost unchanged on a year earlier[4]. Neither number tells you what a buyer paid for the house next door.
Source: Propertymark, 2026[2]
Show the data
| Item | Value |
|---|---|
| Below asking | 86% |
| At asking | 9% |
Use asking prices for one job only: to show the seller what they will be competing against. Use completed sales from the price paid data for the figure itself.
A six-step method for pulling comparables
Set the search area
Start with the same street or the same estate, then widen to streets of the same age and character within walking distance. Stop widening once you have enough evidence.
Match type and tenure
Filter by property type and by freehold or leasehold. A leasehold flat is not evidence for a freehold house, and a short lease changes value on its own.
Choose the time window
Begin with the last 12 months of completed sales. Go back further only for unusual homes, and adjust older sales for the market move.
Confirm size and layout
The data has no floor area or bedroom count, so check the old listing, the EPC floor area or your own records for each sale you keep.
Remove the noise
Drop new builds unless you are valuing one, Category B entries and any sale you know was between family members.
Keep three to six
A short list of strong comparables persuades a seller more than twenty weak ones. Note why each was kept.
Adjusting for time, size and condition
Every comparable needs a written adjustment, even if the adjustment is nil. Time is the first. Sales complete months after the price is agreed, and the government puts the average time from accepted offer to completion at 120 days[5]. A sale that completed in March was probably agreed before Christmas, so compare the agreed date to today's market, not the completion date.
| Comparable | Recorded price | Time | Size and layout | Condition | Adjusted view |
|---|---|---|---|---|---|
| 14 Elm Road, semi, 11 months ago | 312,000 | Flat market, nil | Same layout, nil | Dated kitchen, add | Slightly above recorded |
| 3 Ash Close, semi, 4 months ago | 325,000 | Nil | Extra bedroom, deduct | Similar, nil | Below recorded |
| 27 Elm Road, semi, 7 months ago | 318,000 | Nil | Same layout, nil | Recently refitted, deduct | Below recorded |
Size is next. Where you have floor areas, compare price per square metre across the comparables, then ask whether the subject's layout earns or loses value against that rate. Condition comes last and is the one only you can see: the data records nothing about kitchens, roofs or damp.
Example: a three-bedroom semi in a market town
Example: a negotiator values a 1960s three-bedroom semi. The price paid data shows eleven sales on the estate in 18 months. She drops two new-build plots, one Category B repossession and a transfer she knows was between siblings, leaving seven. Four are the same house type within a year. Two of those were extended, so she keeps them but notes a deduction. Her recommended range rests on four sales, and she can name the address behind every pound.
Example: the seller has seen a similar house down the road on the market for much more. The negotiator shows the asking price next to the completed sales and explains that most homes are currently selling under asking. The conversation moves from her opinion against his hope to evidence both can read.
When the evidence is thin
Some homes have no clean comparables: the only thatched cottage in the village, a converted chapel, a house on a plot three times the size of its neighbours. Do not stretch weak evidence to look strong. Say plainly that the evidence is limited and show how you built the figure instead.
- Widen the time window and adjust each older sale for the market move since, stating the basis for the adjustment.
- Widen the area to villages or streets with a similar buyer profile, and say why you think buyers would treat them as alternatives.
- Use the nearest standard house type as a floor and explain what the unusual features add or take away.
- Ask a colleague who has sold similar homes to review the range before it goes to the seller.
- Recommend a pricing format that lets the market set the level, and agree a review date after the first viewings.
A seller with an unusual home usually knows it is hard to value. Being open about the gaps earns more trust than a precise figure that falls apart on the first viewing.
Checking flags before you rely on a sale
Three fields in each record deserve a second look before a sale goes on your sheet. The new-build flag matters because first sales of new homes often include incentives, so they rarely match what a second owner achieves. The tenure field separates freehold from leasehold, and a mixed set of flats and houses on one estate can produce a misleading average. The category field separates standard sales from repossessions and other additional entries[1]. Spend a minute on each field and you avoid most of the errors that sellers spot first.
Showing the evidence to the seller
- Put the comparables on one page: address, date, price, type, tenure and a one-line adjustment note.
- Show the street on a map so the seller sees you did not cherry-pick distant sales.
- Say which sales you rejected and why. Sellers often know the street better than you and will ask.
- Give a range and a recommended marketing figure, and explain how the two relate. See asking price strategy for the format choice.
- Flag the two-month lag so the seller is not surprised when a recent sale appears later.
Keep the working on file
The Property Ombudsman's code for residential estate agents says any figure you advise as an asking or selling price must be given in good faith, reflect current market conditions and be supported by comparable evidence, and that the evidence must be kept on file[3]. Save the adjustment sheet with the valuation record. If the seller later complains that the figure was inflated or too low, the file shows how you reached it.
Regional averages are useful context but not comparables. Average prices in the year to August 2026 were £294,000 in England and £216,000 in Wales[6]; a valuation built on either would be meaningless for a single street.
Frequently asked questions
How up to date is the Land Registry price paid data?
It is published monthly, but registration lags the sale, so the two most recent months are incomplete. Treat a lack of recent sales on a street with caution.
Does the price paid data include Scotland?
No. It covers residential sales in England and Wales lodged with HM Land Registry. Use other sources for Scotland and Northern Ireland.
Why can't I find a sale I know happened?
It may not be lodged yet, or it may be excluded: gifts, right to buy discounts, transfers of a share between divorcing couples and court-ordered transfers are all left out.
How many comparables should I show a seller?
Three to six strong ones, matched on type, tenure and area, each with a written adjustment, usually persuade better than a long list.
References
HM Land Registry (GOV.UK) (2026). About the Price Paid Data (opens in a new tab) ↑
Propertymark (2026). Housing Insight Report, August 2026 (PDF) (opens in a new tab) ↑
The Property Ombudsman (2026). Code of Practice for Residential Estate Agents (TPOE27-9, March 2026) (opens in a new tab) ↑
Rightmove (2026). House Price Index, October 2026 (opens in a new tab) ↑
Ministry of Housing, Communities and Local Government (GOV.UK) (2025). Home buying and selling reform (consultation) (opens in a new tab) ↑
Office for National Statistics (2026). Private rent and house prices, UK: October 2026 (opens in a new tab) ↑