How UK Property Valuations Work

By Sean R. · Updated 18 August 2026

"What's my house worth?" gets answered at least three different ways in the UK property market, and it's worth knowing how each one arrives at a number.

1. Automated valuation models (AVMs)

Tools like this one work by finding recent sales of comparable properties near yours, adjusting each one to today's prices using a house price index (since a sale from three years ago needs scaling up or down to reflect how the whole market has moved since), and taking the median. The result comes with a confidence score based on how many comparables were found, how close by, how recent, and how much they agree with each other.

What an AVM can't see: the inside of your property. It doesn't know about your extension, your damp problem, or your new kitchen, it only knows what similar properties nearby have sold for.

2. Estate agent appraisals

An agent valuing your home for sale combines the same comparable-sales logic with local market knowledge (what's selling right now, not just what sold historically), and a physical look at your property's condition and any improvements. Agents also have a commercial incentive to give an optimistic figure to win your listing, worth bearing in mind when comparing an agent's appraisal to an AVM's more dispassionate number.

3. RICS surveys and mortgage valuations

A formal valuation by a RICS-qualified surveyor is the most rigorous: a physical inspection, professional judgement on condition and any defects, and a valuation that a mortgage lender is willing to lend against. This is what determines whether a sale can complete at the agreed price. See our guide on RICS surveys vs online valuations for more detail.

The common thread: comparable sales

All three methods ultimately lean on the same underlying idea, what have similar properties sold for recently, nearby? The differences are in how rigorously that comparison is done, how much local/physical knowledge is layered on top, and how much of an incentive the person or system giving you the number has to be right versus favourable.

What moves the number

Across all methods, the biggest factors are: location (down to the street, sometimes), property type and size, condition, tenure (see our leasehold vs freehold guide and, for short leases specifically, the short lease value cliff edge), energy efficiency (see our EPC and value guide), and how recently comparable properties have sold. See our full guide on what affects property value for a complete breakdown. If you're budgeting for a purchase, also see our Stamp Duty guide for what to add on top of the valuation itself.

A worked example: how this site reaches a number

Say you enter a postcode for a 3-bedroom semi. This tool first pulls HM Land Registry sold prices for 3-bedroom semis in that postcode sector, going back as far as needed to find enough genuinely comparable sales, starting tight and widening only if the sector is sparse. Each sale is then adjusted using the UK House Price Index for its region and sale date, so a sale from two years ago is expressed in today's terms rather than treated as current. The adjusted sales are checked against a sqft-plausibility gate to exclude any with implausible recorded floor areas for their bedroom count (a rare but real data-entry issue in any large public dataset), then the median of what's left becomes your estimate, alongside a confidence score reflecting how many comparables were found, how close, and how recent. See our full methodology page for the complete detail.

Frequently asked questions

How many comparable sales does this tool use?

As many genuinely comparable sales (same postcode sector, similar bedroom count and property type) as it can find within the tightest reasonable radius and time window, widening only when there aren't enough to produce a reliable figure, the confidence score tells you how much widening was needed.

Why does the estimate come with a range or confidence score instead of just a number?

Because a single headline figure hides how reliable it is. A number built from 60 recent nearby sales deserves far more trust than one built from 4 sales three years ago at the edge of a wide radius, even if both display as a specific-looking figure, the confidence score makes that difference visible instead of hiding it.

How often is the underlying data updated?

HM Land Registry publishes sold-price data on a rolling basis (with an inherent reporting lag between completion and publication), and this site's comparable-sales cache refreshes on a regular schedule to reflect newly published sales as they become available.

Why might this tool's estimate differ from an estate agent's appraisal for the same address?

An agent layers in current local market knowledge and a physical look at your property (plus, often, a commercial incentive to quote optimistically to win your listing), while this tool relies purely on adjusted historical comparable sales. Neither is "more correct" by default, they're answering the same question with different inputs, similar to the gap explained in our Zoopla vs Land Registry guide.

In summary

• All valuation methods, AVMs, agent appraisals, RICS surveys, ultimately rely on the same core idea: what have comparable properties sold for recently, nearby.
• This tool's process: find comparable sales in your postcode sector, HPI-adjust each to today's prices, filter implausible records, then take the median with a confidence score.
• What differs between methods is how much local knowledge, physical inspection, or commercial incentive gets layered on top of that core comparable-sales logic.
• The biggest factors moving any valuation are location, size, condition, tenure, and how recent the comparable sales are, see our full breakdown.

Understanding this process matters most when a valuation surprises you, either higher or lower than expected. Before assuming the figure is wrong, check the confidence score and comparable count behind it: a low-confidence estimate built from a handful of older, wider-radius sales is telling you it's a rougher starting point, not that the methodology has failed.

Related situations worth knowing about

A few specific situations come up often enough to deserve their own guide: if a lender's valuation comes back below your agreed offer, see why mortgage valuations can be lower than your offer. If you're valuing a property for probate or Inheritance Tax, that's a different exercise from a sale valuation, see how to value a property for probate. And if you bought new and a valuation now looks lower than what you paid, see why new-build homes can show a lower value in the first year.