1. Overview

This dataset is a deduplicated register of every formally registered civil society organisation in the United Kingdom. It was produced by the UK Third and Civil Society Sector Database (TCSS) project, which collects, processes, and links public administrative data on civil society organisations across the United Kingdom.

The Organisation Register — referred to as the “Data Spine” — brings together records from ten UK regulatory bodies into a single dataset. Each organisation receives a unique identifier (UID) that enables records to be linked across data sources. A matching algorithm identifies organisations that appear in more than one register and links them under a single UID, producing a comprehensive, deduplicated picture of the sector.

The register is distributed as a set of four CSV files that together describe the organisations, their cross-register links, and their industrial classifications. All data in the register are derived from openly available public sources.

2. What is the Organisation Register?

The UK civil society sector is governed by a patchwork of regulatory bodies. Charities in England and Wales are registered with the Charity Commission for England and Wales (CCEW); charities in Scotland with the Office of the Scottish Charity Regulator (OSCR); and charities in Northern Ireland with the Charity Commission for Northern Ireland (CCNI). Companies — including Community Interest Companies, charitable companies, and other nonprofits limited by guarantee — are registered at Companies House. Co-operatives and mutual societies appear on the Co-operatives UK register and the FCA Mutuals Public Register. Sector-specific bodies such as the Care Quality Commission, Care Inspectorate Scotland, and the social housing regulators in England and Scotland maintain further records.

Many organisations appear in more than one of these registers. A charity incorporated as a company limited by guarantee, for example, will have records at both a charity commission and Companies House. Without linkage, these appear as separate entities, leading to double-counting and fragmented analysis.

The Organisation Register solves this problem by:

• Ingesting data from all ten source registers
• Normalising names, addresses, and dates into a common format
• Running a matching algorithm to identify duplicate records across registers
• Assigning a single UID to each unique organisation
• Storing primary records in the spine file and alternative records in the supplementary file

The result is the first comprehensive, deduplicated listing of UK civil society organisations available as open data. Further details of the construction methodology are available in the Spine Builder documentation and the project’s GitHub repository.

3. Dataset Contents

The register is distributed as four CSV files:

Organisation Types

The cso_type field classifies organisations into four main types: Charity, CIC, Co-operative/Mutual, and Other. The cso_subtype field provides a finer-grained classification: Charities, CICs, and Co-operatives/Mutuals pass through unchanged, while the ‘Other’ category is disaggregated into subtypes using Companies House SIC (Standard Industrial Classification) codes and organisation name patterns. The classification applies SIC division codes (the first two digits of the primary SIC code) alongside keyword matching on organisation names. Rules are applied in a documented first-match priority order: each organisation receives the first subtype whose rule it satisfies (dormant companies first, then SIC-division rules, then name-pattern rules). Name-pattern rules also apply when an organisation has no usable five-digit SIC code, so an organisation cannot escape classification simply because its SIC information is missing or in a legacy format.

Table 1 shows the composition of the sector by organisation type, country, and status. Percentages are calculated within each status-country group (for example, of all active organisations in Scotland, 78.3% are charities).

Property management companies dominate the ‘Other’ category among active organisations in England (44.0% of active ‘Other’ organisations, equivalent to about 15% of all active English organisations) and are also prominent in Wales (22.3%), reflecting the prevalence of residential leasehold structures; they are far rarer in Scotland (3.0%). The residual ‘Other Company Limited By Guarantee’ category remains the largest subtype among dissolved ‘Other’ organisations in every country (59–66%).

4. Coverage & Composition

The register covers 782,995 organisations, of which 385,515 (49.2%) are currently active (no removal date) and 397,480 (50.8%) are dissolved or removed.

Organisations by Source Register

Community Interest Companies

Of the 782,995 organisations in the register, 78,595 (10.0%) are flagged as Community Interest Companies. The is_cic field provides a direct way to filter for this subset.

Companies House Company Types

Organisations sourced from Companies House are filtered to include only civil-society-relevant company types. The included types and their approximate counts are:

Excluded company types include Private Limited Companies, Public Limited Companies, Limited Partnerships, and other company forms that are not typically associated with the civil society sector. The full inclusion and exclusion criteria are listed in Section A4.

5. Unique Identifiers

Each organisation in the spine is assigned a UID that encodes its source register and original registration number. The UID format follows the org-id.guide convention where possible.

UID Prefix Scheme

6. Data Sources

The register draws on open data from ten UK regulatory bodies. The table below lists each source and its URL.

The first eight sources in the table above contribute primary records to the spine. The Care Inspectorate Scotland and Care Quality Commission contribute only to the matches file — they are used to link organisations to existing spine records but do not add new organisations to the register.

7. Limitations & Caveats

• Scope. The register covers formally registered organisations only. Unregistered voluntary groups, informal associations, and community groups that do not appear on any regulatory register are not included.
• Coverage by jurisdiction. Coverage varies between jurisdictions. The CCNI register was established more recently than its counterparts in England, Wales, and Scotland, so the Northern Ireland data has less historical depth. CCNI also does not provide removal dates for deregistered charities.
• Missing data. Some fields are incomplete. Approximately 13,200 CCEW charity records lack address information (most were removed before 1997). Around 1,800 OSCR records lack organisation names and addresses. The city field is frequently blank across all sources.
• Multiple dates. A small proportion of CCEW records (3.6%) have more than one registration date. This can arise from data entry corrections, organisations that were removed and re-registered, or merged charities. The spine records the earliest registration date; the removal date follows the newest authoritative source record (see the Current status note below), with other date variants retained in the supplementary file.
• Matching limitations. The deduplication algorithm identifies most cross-register duplicates but cannot guarantee perfect recall. Some genuine duplicates may remain unlinked (false negatives), particularly where organisation names differ substantially between registers or where no company number link exists. Conversely, a small number of false positive matches are possible where two genuinely distinct organisations share a normalised name.
• Point-in-time snapshot. The register reflects data as available at the time of construction. Organisations registered or removed after the data download dates will not be captured. The current release incorporates data downloaded through July 2026, plus historical register content carried forward from the previous release (see the changelog).
• Postcodes and geography. To link organisations to geographic areas (Local Authority, region, country), the postcode field can be joined to the National Statistics Postcode Lookup (NSPL). However, not all organisations have a valid postcode in the register.

8. What Can You Learn?

The Organisation Register enables a range of research and policy applications:

• Sampling frame for national surveys. The deduplicated register provides the first comprehensive sampling frame that crosses legal form and organisation type within a single dataset. Researchers can draw stratified samples by organisational type, jurisdiction, geography, industrial classification, and organisational age.
• Local ecosystem analysis. By linking organisations to Local Authority areas via postcodes, the register enables analysis of how civil society provision varies across communities. Per-capita measures can be calculated by joining population estimates from the ONS mid-year estimates.
• Longitudinal tracking. The combination of registration dates, removal dates, and the linked financial datasets enables tracking of organisational lifecycles — from incorporation through growth, decline, and dissolution.
• Cross-register linkage. The UID system and matches file enable researchers to link the Organisation Register with other TCSS datasets, including charity financial records, Companies House accounts, CIC 36 community interest statements, CIC board of directors data, and public procurement records.
• Sector mapping. The SIC codes file enables industrial classification of Companies House organisations. Combined with the ICNPTSO charity classifications (available separately via the UK Charity Activity Tags project), researchers can map the full breadth of civil society activity.

9. Citation & Licence

Suggested Citation

McDonnell et al. (2026). UK Civil Society Organisation Register. UK Third and Civil Society Sector Database. Available at: https://uk-third-sector-database.github.io/data/. Licensed under CC BY 4.0.

If you would like to learn more about this dataset and how it can be applied to your project or research programme, please contact research@brawdata.com.

10. Changelog

A1. Data Processing

Each source register uses different conventions for field names, encodings, date formats, and the treatment of alternative names and addresses. The pre-processing step for each source:

1. Identifies the primary name and address for each organisation (choosing the most recent where multiple iterations exist)
2. Stores alternative names, addresses, and dates in the supplementary file
3. Normalises names by removing leading/trailing spaces, collapsing consecutive internal spaces, and converting to uppercase
4. Normalises addresses by concatenating address lines and the postcode into a single string
5. Converts all dates to dd/mm/yyyy format

The eight primary sources are then combined into a single dataset. Organisations already present from a higher-precedence source are not duplicated; instead, any additional data is stored in the supplementary file. The project source code is available on GitHub.

A2. Source Precedence

When an organisation appears in multiple registers, the register with the highest precedence supplies the primary record (UID, name, address, dates) for the spine file. The precedence order is:

Charity registers take precedence over Companies House, meaning that a charity with a Companies House registration will have a GB-CHC- or GB-SC- UID rather than a GB-COH- UID. The Companies House record will appear in the supplementary and matches files.

A3. Matching Algorithm

The deduplication process uses several strategies to identify organisations that appear in more than one register:

Matching by Company ID

The charity regulators (CCEW and OSCR) record the Companies House company number for charities that are also registered companies. This enables deterministic matching:

• CCEW charity records are matched to Companies House records via the company number
• OSCR records with a company identifier are matched to Companies House records
• Co-operatives UK records with a “Registered Number” are matched to the Mutuals Public Register
• Where two CCEW charities share the same company number, they are linked as duplicates

Matching by Normalised Name

Where two organisations in different registers share the same normalised name, they are tentatively matched subject to an additional condition. The match is confirmed if any of the following hold:

1. The organisation is in CQC data and the matched core organisation is flagged as having a CQC registration in CCEW data
2. The organisation is in the OSCR register and has a “cross border” flag indicating a match is expected with CCEW
3. The organisation is in the Scottish Housing Regulator, and the normalised name matches an OSCR or Mutuals Public Register record; or it is in Care Inspectorate Scotland and the normalised name matches an OSCR record
4. The organisation is in Social Housing England, and the normalised name matches a CCEW or Mutuals Public Register record; or it is in CQC and the normalised name matches a CCEW record

Matching via Find That Charity

A lookup table provided by Find That Charity (David Kane) is used to generate additional matches. These links include historic re-registrations, allowing the dataset to be further deduplicated. This is the most common match type, accounting for 73,315 of the 136,578 total links.

Determining the Primary Organisation

Where a match occurs between organisations from different sources, the higher-precedence source provides the primary record. Where a match occurs between organisations from the same source:

• For CCEW matches, the Find That Charity lookup identifies the “transferee” in a re-registration; the transferee is treated as primary
• Otherwise, the organisation without a removal date (i.e. still active) is primary; if both have removal dates, the one with the most recent removal date is primary
• If neither has a removal date, the organisation with the earliest registration date is primary

A4. Companies House Inclusion Criteria

Companies House records are filtered to include only company types relevant to the civil society sector. The following table lists all company categories and whether they are included or excluded from the register.

Included Company Types

Excluded Company Types

A5. Linking to Other TCSS Datasets

The Organisation Register serves as the backbone for the wider TCSS data ecosystem. The uid field enables direct linkage to the following companion datasets: