TL;DR
UK open spatial data is more usable than it was two years ago, driven by the National Data Library, NUAR, the Planning Data Platform, and HM Land Registry's own digitisation programme. But the detailed layers that serious commercial work depends on, OS MasterMap, full title plans, are still licensed, and integration across datasets is still the hard part.
The UK has a genuine wealth of open spatial data. Government agencies, Ordnance Survey, the Environment Agency, Natural England, Historic England, and local authorities all publish datasets that anyone can download and use.
But “open” doesn’t mean “ready to use.” The quality, format, and completeness of these datasets vary enormously. Knowing what’s available is useful. Knowing what’s actually fit for purpose is more useful.
This picture is also moving faster than it has in years. Government is actively trying to fix the fragmentation problem that has defined UK open data for a decade, and 2026 is turning out to be a genuine inflection point.
What’s good
Environment Agency flood data. Flood zones 2 and 3, surface water flood risk, and reservoir flood maps are all published as spatial datasets. Well-maintained, regularly updated, and essential for any site assessment or development appraisal work. This is open data done properly.
Natural England designations. SSSIs, Special Areas of Conservation, National Nature Reserves, Ancient Woodland. Published as polygon datasets with clear metadata. Reliable and current.
Historic England. Listed buildings, scheduled monuments, registered parks and gardens, registered battlefields. Good spatial coverage, well-documented, and freely available.
OS Open Data. Ordnance Survey publishes a range of datasets under open licence: OS Open Roads, OS Open Rivers, Boundary-Line (administrative boundaries), and others. Not the same level of detail as premium OS products, but useful and free.
UPRN data. The full UPRN (Unique Property Reference Number) register is now open. This is significant. UPRNs are the closest thing the UK has to a universal property identifier, and having them publicly available makes data integration substantially easier.
What’s useful but awkward
Energy Performance Certificates. The EPC register is publicly searchable and downloadable. The data is valuable for property analysis, but it’s published in CSV format with addresses rather than coordinates. Geocoding it reliably requires work.
Planning application data. Local authorities are required to publish planning registers, but there’s no national standard. Some publish spatial data. Some publish spreadsheets. Some publish PDFs. Aggregating planning data across the UK is a major data engineering exercise, not a download, though that’s changing, more on this below.
Land Registry. Price paid data and title boundary data (INSPIRE polygons) are both available. The INSPIRE polygons are useful but low-resolution. Detailed title plans require a paid search, for now, HMLR’s own digitisation programme is aimed squarely at this gap (see below).
What’s missing
A national building polygon dataset under open licence. OS MasterMap Topography Layer contains the definitive building footprints, but it’s a premium licensed product. The open alternatives are either lower resolution or incomplete.
Consistent local authority data. Every local authority publishes data differently. Different schemas, different formats, different update frequencies. There’s no national standard for local spatial data publication, and it shows.
Cross-dataset identifiers. Datasets that describe the same places often don’t share identifiers. A building polygon in one dataset, a flood zone in another, a planning constraint in a third. Linking them requires spatial joins rather than simple key lookups.
Current land use data. Comprehensive, current land use classification for the UK doesn’t exist as an open dataset. You can approximate it by combining several sources, but it’s not straightforward.
What you still pay for
Open data covers a lot of ground, but the detailed, high-currency layers that underpin serious commercial work are still licensed.
OS MasterMap and the OS National Geographic Database. Building footprints, land parcels, and topography at the resolution planning and infrastructure work actually needs. This data now lives in OS NGD, Ordnance Survey’s restructured authoritative data store, but access still sits behind the Public Sector Geospatial Agreement (PSGA) for public bodies or a commercial subscription for everyone else. The PSGA is worth roughly £1 billion of public investment, which tells you how central this data still is even as the open layer expands (more on OS NGD below).
Detailed title plans. Land Registry’s INSPIRE polygons are free and low-resolution. The actual title plan, the one that matters for a real transaction, is a paid search.
Commercial data brokers. Firms like Digimap and Emapsite exist precisely because aggregating, licensing, and delivering OS and third-party datasets in a usable form is still a paid service, even where the underlying data has an open equivalent.
Address data. The Royal Mail Postcode Address File has historically sat behind a licence. That’s now changing (more below), but plenty of address-matching workflows are still built on paid AddressBase products.
None of this is a criticism. Premium data exists because someone has to survey, verify, and maintain it to a standard that free datasets don’t attempt. The point is that “open data” and “all the data you need” are not the same claim, and budgeting for the paid layer is normal, not a failure of research.
The bigger shift: aggregation is finally happening
For years, the honest complaint about UK open data was that it existed but was scattered. Every local authority published differently. National datasets and local datasets didn’t share identifiers. Finding the right dataset was often harder than using it once found. That’s the backdrop this piece was written against.
2026 is the year that started to change, on several fronts at once.
The National Data Library. DSIT is turning data.gov.uk from a largely uncurated directory into what it calls a single trusted gateway to curated public sector data, backed by more than £100 million of the government’s wider £1.9 billion digital investment. Five pilot projects are underway: linking data for energy bill support, cutting admin burden for people with long-term health conditions, improving adult social care data access, making legal records AI-ready for small business guidance, and widening use of Met Office climate data. A January 2026 progress update confirmed the “discovery phase” is done, with more detail promised for spring 2026. The roadmap is explicit that this is a slow build: current work is about curated collections and fixing the existing directory, with search, discovery, and platform redesign listed as “next,” and deeper interoperability as “later.” It’s also explicit about what it won’t do: no mandated standards, no data exchange tooling, no hosting service. Worth knowing before you plan around it.
The National Underground Asset Register (NUAR). Operated by Ordnance Survey, NUAR is now genuinely operational: over 350 asset owners contributing data, more than 3 million km of mapped pipes and cables out of roughly 4 million km of underground pipes and cables in the UK. Government puts the economic case at over £400 million a year in reduced accidental strikes, which currently run at around 60,000 a year and cost roughly £2.4 billion annually. Access today is restricted to asset owners, statutory undertakers, and their authorised contractors, not the general public, but a spring 2026 consultation is looking at extending access to cases like EV chargepoint rollout, flood risk planning, and property development. This is a genuinely new category of national dataset that didn’t meaningfully exist five years ago, even if it isn’t open in the download-it-yourself sense.
Planning data. MHCLG’s Planning Data Platform is trying to turn England’s fragmented local planning registers into one consistent resource, built around explicit data standards (right down to enforcing one date format so nobody has to guess whether 1-2-2020 means January or February). New regulations now require local planning authorities to publish plan and timetable data in a standardised format, the first real use of the data-standardisation powers in the Levelling Up and Regeneration Act 2023. MHCLG has also rolled out Extract, a free AI tool built with i.AI that converts scanned planning documents, conservation area maps, Article 4 Directions, Tree Preservation Orders, into structured data, cutting a job that used to take an officer up to two hours down to about two minutes. It’s now available to every local planning authority in England. This is the planning data problem described earlier in this piece, being worked on at scale, in public, with the gaps still visible.
HM Land Registry’s own digitisation push. The most directly relevant shift for anyone working with property or land data: HMLR is working toward what it calls a fully digital, geospatial land register, combining ownership, leasehold, rights of way, and environmental data on interactive maps, rather than the current system of static title plans and paper-derived records. To help deliver it, HMLR has appointed CGI as its Strategic Data Partner, a contract worth up to £14.1 million running from July 2026, covering geospatial strategy, data governance, and data integration across HMLR’s transformation programme. Near-term work for 2026/27 includes using AI to convert historic, handwritten legal documents, indentures, into structured digital data. None of this is finished, and a geospatial register at national scale is a multi-year undertaking, but it’s the clearest sign yet that the “paid search for a title plan” model described above has a shelf life.
OS National Geographic Database (OS NGD). This is the paid side of the shift, not the open side, and worth including precisely because the two are easy to conflate. OS has rebuilt its entire data model around one authoritative store of over half a billion geographic features, organised into 16 themed collections, replacing the old model of downloading whole separate legacy products. It’s still licensed, through the PSGA for public sector bodies or a commercial subscription for everyone else, but the consistency it brings matters for anyone integrating OS data with other national datasets: shared structure across collections, daily updates, and a Royal Mail Address Feature Type, added in September 2025, that folds the Postcode Address File into OS’s geospatial attribution. The premium layer is getting easier to work with even as it stays behind a paywall.
UK Geospatial Strategy 2030. The Geospatial Commission’s refreshed strategy sets national priorities through the decade and includes a review of the PSGA itself, the roughly £1 billion agreement that funds public sector access to OS data. Expect the boundary between “open” and “public sector licensed” to keep shifting as that review lands.
Taken together, these aren’t isolated projects. They’re a coordinated attempt to fix the exact problems this piece describes: fragmentation, inconsistent local data, and the absence of shared identifiers across datasets. None of it is finished, and some of it (NUAR, OS NGD, HMLR’s register) is really about making licensed or restricted data more usable rather than making it free. But the direction is clear, and it’s worth watching if you plan on this data for more than a one-off project.
The practical takeaway
UK open spatial data is genuinely useful, but rarely sufficient on its own, and that’s less true today than it was two years ago. Most real-world projects still combine open data with licensed datasets, client data, and derived datasets. The open data provides the foundation. The value comes from integration, and integration is exactly what the current wave of government initiatives is trying to make easier.
The identifier and integration problems described above aren’t abstract. We ran into the same fragmentation, no shared keys, inconsistent formats across local authorities, building a national building intelligence platform from 44 million UK building polygons and planning data spanning 317 local planning authorities.
If you’re building a platform or workflow that depends on UK spatial data and want to understand what’s available now, what’s still worth paying for, and what gaps need filling, our property and land GIS support is a good place to start, or get in touch directly.