Published 21 May 2026

UK postal address validation explained

Last updated: 21 May 2026

UK postal address validation matches each address against the Royal Mail Postcode Address File (PAF), the authoritative file of around 31 million UK delivery points. Validation does three things: confirms the address exists, corrects format and capitalisation to PAF standard, and adds the Royal Mail delivery point suffix (DPS) for Mailmark eligibility. PAF licence is required (typically £200 to £2,000 per year depending on volume). Validation typically corrects 5% to 15% of unvalidated records and rejects 1% to 3% as undeliverable.

Key points

What is the Royal Mail Postcode Address File?

The Postcode Address File, universally abbreviated to PAF, is the single authoritative source of postal addresses in the United Kingdom. Royal Mail maintains it under its statutory obligations as the UK's universal service provider. The file holds approximately 31 million delivery points: every residential property, commercial premises, and PO Box that Royal Mail delivers to. It is updated weekly, typically adding new builds within weeks of Royal Mail assigning a postcode, and removing demolished or decommissioned addresses on a rolling basis.

PAF is not the same as the Code-Point Open dataset (which maps postcodes to grid references) or the Ordnance Survey AddressBase Premium file (which adds property-level GIS coordinates). For postal marketing purposes, PAF is the correct reference file because it reflects actual mail delivery. Code-Point Open tells you where a postcode sits geographically; PAF tells you whether a specific letter can be delivered there.

Access to PAF requires a licence from Royal Mail. Direct licences are available to businesses but carry usage conditions including restrictions on redistribution and sub-licensing. Annual fees typically range from around £200 for low-volume applications to £2,000 or more for high-volume commercial users. In practice, the majority of UK marketers never licence PAF directly: they pass their files through a bureau or API service that holds the licence and returns validated results.

PAF licence: bureau route vs. direct licence

For a one-off cleanse of 50,000 records or fewer, a bureau service charging per-record is almost always cheaper than acquiring a direct PAF licence. Direct licences become cost-effective for organisations running high volumes of real-time lookups, such as e-commerce checkouts or CRM import workflows, where per-record bureau fees would accumulate quickly.

What are the three outputs of postal address validation?

Asking a validation service to check an address produces three distinct outputs. Understanding the difference matters because each output has a separate downstream effect on your campaign.

Output 1: existence confirmation

The first and most fundamental output is a binary pass or fail. The validation engine checks whether the submitted address, after normalising whitespace and common abbreviations, corresponds to a delivery point in PAF. A record that fails this check is marked as undeliverable. Royal Mail will not attempt delivery to an address that does not exist in its system, so any piece of mail sent to that record is wasted postage.

In our experience, the rejection rate on a consumer database that has never been validated sits between 1% and 3%. On older B2B files with high staff turnover and premise churn, rejections can reach 5% or more. At 50p per direct mail piece, a 2% rejection rate on a 100,000-record file represents £1,000 in immediate avoidable waste, before you count production costs.

Output 2: format correction to PAF standard

PAF enforces a specific address layout: up to seven address lines with defined precedence rules for building name, sub-building name, organisation name, thoroughfare, dependent thoroughfare, post town, and postcode. Most address records gathered from web forms, CRM imports, or third-party feeds do not conform to this layout. Common errors include:

Correcting these errors matters beyond aesthetics. Royal Mail's Mailsort and Cleanmail presort schemes require address data to meet PAF-standard formatting before you qualify for bulk-mail discounts. A file that is 95% matched to PAF but formatted inconsistently may still fail pre-sort qualification.

Output 3: the delivery point suffix (DPS)

The delivery point suffix (DPS) is an eight-character alphanumeric code that uniquely identifies a single delivery point within a postcode. PAF validation appends the DPS to each matched record. When combined with the postcode, the DPS enables Royal Mail's Mailmark barcode scheme, which offers the lowest available postage tariffs on qualifying bulk mailings.

Without the DPS, your mail can still be delivered, but you cannot access Mailmark pricing. For a high-volume direct mail programme, the Mailmark discount can reduce postage by several pence per item. On a 500,000-piece mailing that is material saving. The DPS is also used internally by Royal Mail's automated letter-sorting equipment, so addresses with a valid DPS are processed faster through the network.

Validation outputs at a glance

Output What it checks What it fixes or adds Why it matters
Existence check Does this address exist as a live PAF delivery point? Flags undeliverable records for suppression Eliminates wasted postage on undeliverable pieces
Format correction Does the address layout match PAF conventions? Corrects capitalisation, abbreviations, line order, postcode spacing Required for Mailsort/Cleanmail presort qualification
DPS append Is a delivery point suffix available for this record? Appends the eight-character DPS code Unlocks Royal Mail Mailmark postage discounts

Why do addresses fail validation?

An address failing the PAF existence check does not automatically mean it was fabricated or entered in bad faith. Several legitimate scenarios produce a genuine property that PAF does not yet hold or no longer holds.

New builds

Royal Mail adds new residential and commercial properties to PAF on a rolling weekly basis, but there is always a lag between a property being physically completed and its PAF record being live. A housebuilder completing 200 units in June may not see all addresses in PAF until late July. Consumer databases sourced before completion will contain new-build addresses that temporarily fail validation. The correct response is not suppression but a re-validation pass three to four weeks later, by which time the PAF record will typically be present.

Recently demolished properties

Conversely, Royal Mail removes addresses when properties are demolished or when postcodes are reorganised following significant redevelopment. A record that validated cleanly twelve months ago may now return a non-match if the building has since been cleared. This is one of several reasons postal validation is not a one-time exercise: files decay.

BFPO addresses

British Forces Post Office (BFPO) addresses, used by serving armed forces personnel and their families, are held on a separate secure register managed by the MoD and are not available in the standard PAF file. A correctly formatted BFPO address will fail standard PAF validation. If your file includes BFPO records, they should be routed through a separate validation process or held out from the PAF pass entirely.

Business premise changes

Commercial properties change status frequently. A warehouse converted to residential loft apartments may hold records formatted as "Unit 3, Blackfriars Industrial Estate" in your CRM while PAF now records individual flat numbers. The old business address format will fail; the new residential format must be obtained from the occupant or, for B2B purposes, treated as lapsed.

How validation fits into the broader postal cleansing workflow

PAF validation is one step in a postal data quality chain. It is necessary but not sufficient on its own. A fully validated address confirms that the delivery point exists and is correctly formatted. It says nothing about whether the person you want to reach still lives or works there.

That is why validation pairs with NCOA (National Change of Address) processing. The order matters. Run PAF validation first to standardise your addresses to PAF format; NCOA matching uses PAF-format addresses as its input, so unvalidated records with inconsistent formatting will generate false non-matches. Once your file is PAF-clean, NCOA can reliably identify records where the individual has registered a move with Royal Mail's Redirection service and return their current address.

If you are building a file for direct mail from scratch rather than cleaning an existing database, the same principle applies. Postal append services add addresses to records that arrive with only a name or email, sourcing against opt-in consumer data. The appended addresses should already be PAF-validated at point of supply, but a secondary validation pass before mailing is good practice, particularly if there is more than a four-week gap between data supply and print despatch.

Validation is not a one-off task

UK addresses decay at around 8% per year across a mixed residential and commercial file, driven by house moves, demolitions, and business closures. A validation pass carried out twelve months ago offers only partial protection. For files used in ongoing direct mail programmes, best practice is a re-validation pass at least once per year, or before any mailing with a gap of more than six months since the previous cleanse.

How accurate is PAF validation, and what can it not do?

PAF validation is highly accurate for what it is designed to check: whether an address exists as a Royal Mail delivery point and whether it is formatted to PAF standard. The file itself is maintained by Royal Mail to a very high standard; weekly updates mean lag times for new builds and demolitions are typically measured in days to weeks rather than months.

Typical outcomes across a mixed consumer or B2B file:

What PAF validation cannot tell you is whether the named individual still occupies that address. A perfectly formatted, PAF-matched address to a property where the target moved out two years ago will pass validation without comment. That is why occupancy currency requires NCOA, not just PAF validation. The two processes answer different questions: "does this address exist?" versus "does this person still live there?"

PAF also does not validate the name component of the record. An address like "1 High Street, Exeter, EX1 1AA" will match PAF regardless of whether the name field contains "John Smith" or a sequence of random characters. Name validation is a separate enrichment process involving electoral roll matching or similar consumer-identity files.

Need GDPR-compliant data for your next campaign?

Tell us your targeting criteria and we will run a free count. B2B decision-makers, B2C consumer files, or CRM enrichment, all live verified.

Request Data Counts

Frequently asked questions

What is the Royal Mail Postcode Address File (PAF)?
The Postcode Address File (PAF) is Royal Mail's authoritative database of all UK delivery points, covering approximately 31 million addresses including residential, business, and BFPO records. It is updated weekly and is the reference file used by all accredited address-validation services in the UK. A PAF licence is required to access it, and annual fees typically range from around £200 for low-volume users to £2,000 or more for high-volume commercial applications.
What are the three outputs of UK postal address validation?
Validation against PAF produces three outputs: (1) an existence check confirming the address is a live, known delivery point; (2) a format correction bringing address lines, capitalisation, and postcode into PAF-standard layout; and (3) the delivery point suffix (DPS), an eight-character code that identifies the precise delivery point within a postcode and is required for Royal Mail Mailmark discounts.
How accurate is UK postal address validation?
Validation typically corrects 5% to 15% of records in an unvalidated file, most often fixing capitalisation, abbreviated street types, or transposed postcode characters. Between 1% and 3% of records in a typical consumer or B2B database will be rejected as undeliverable because they do not match any PAF entry. Accuracy depends on the age of your data; files older than twelve months tend to sit at the higher end of those rejection ranges as properties are demolished or postcodes reorganised.
Why might a valid-seeming address fail PAF validation?
Common failure causes include: new-build properties not yet added to PAF (Royal Mail adds new builds on a rolling basis, usually within weeks of Royal Mail assigning a postcode, but there is always a lag); recently demolished properties whose PAF records have been removed; BFPO (British Forces Post Office) addresses, which are held on a separate secure register and are not in the standard PAF file; and former business premises that have been converted to residential use or vice versa where the address format no longer matches the PAF record.
Do I need a PAF licence to validate addresses?
Yes. Direct access to the PAF data requires a licence from Royal Mail, typically priced from around £200 per year for low-volume use to £2,000 or more per year for high-volume commercial applications. Most organisations use a third-party bureau or SaaS validation service that holds the PAF licence on your behalf, so you pay for validation as a service rather than licensing the raw file directly.
How does postal address validation relate to NCOA?
Validation and NCOA (National Change of Address) are complementary but distinct processes. Validation confirms that a postal address exists and formats it to PAF standard. NCOA checks whether the person at that address has moved and, if so, provides their new address. Best practice is to run validation first to clean and standardise your file, then pass it through NCOA to catch movers. Running them in the wrong order means NCOA may fail to match records whose formatting does not align with its reference file.