Short answer
The indexed establishes two principal update schedules:
- nine Business Register entity-data packages are described as updated once a day; and
- the separate annual-report data series is described as initially updated once a month.
The daily files are current-register snapshots, not a complete historical archive. The catalogue expressly says that they contain public data only for legal entities currently in the Business Register with one of three statuses: entered in the register, in liquidation or in bankruptcy. Deleted entities are outside that stated file scope.
Some daily packages contain historical-looking fields—first-entry dates, registry-card entries and dates, rulings, legal succession, previous annual reports or officeholder start and end dates. Those fields can supply a partial event history for an entity that remains in the current snapshot. They do not make the files a complete record of every entity that formerly existed or every previous version of each field.
The catalogue also describes two narrower forms of history:
- annual-report files contain submission metadata and selected indicators for multiple reporting years; and
- the public statistics page permits aggregate queries by year or month, including registrations, deletions and bankruptcy proceedings.
Neither is a row-level historical company register. The annual-report record is also internally inconsistent: it says “the three previous financial years” while listing indicator files for 2019–2025. That conflict is tracked in .
“Daily” describes publication cadence, not a complete history
The verified , and catalogue records each state “Update frequency: daily.” The same daily statement is repeated for the other Business Register bulk packages.
This means RIK says it refreshes the downloadable file once a day. It does not mean:
- the data is real-time;
- every fact about every company is reverified each day;
- every daily version remains available from the source;
- the file includes entities deleted from the register; or
- a current download can reconstruct a company's value for every past date.
A user must retain both the source-file date and the date of retrieval. If an organisation needs its own change history, it must lawfully preserve successive snapshots or obtain an appropriate historical service rather than expecting the newest daily ZIP to contain every prior state.
The daily bulk files cover current-register entities only
The catalogue's repeated coverage warning is unusually specific. The open-data files contain public data only for legal entities currently in the Business Register and only with these statuses:
- entered into the register;
- in liquidation; or
- in bankruptcy.
Deleted entities are not included in that stated current-file population. Failure to find a registry code in today's Basic or General file therefore supports only a limited conclusion: the entity was not found in that file and current-status scope. It does not prove that the entity never existed, never used the name or never had the recorded officeholder, pledge or filing.
The scope warning applies even to datasets that contain event dates. A current file can include historical facts about an entity that remains in scope while omitting an entity that has since been deleted.
Basic data gives current attributes plus one foundational date
The Basic data package contains:
- legal-person name;
- registry code;
- legal form and subtype;
- VAT number;
- current status;
- date of first entry; and
- address.
The date of first entry is historical in the ordinary sense, but it is not a version history. The package does not claim to provide every previous name, address, status transition or registration entry. Its “current status” field reinforces the snapshot character of the file.
A company present in the file can therefore have an old first-entry date and a current address. The catalogue does not establish what its address was on any intermediate date.
General data contains embedded history, but only within the current population
The General data package incorporates Basic data and adds information such as:
- legal succession;
- a list of annual reports;
- a list of articles of association;
- a list of registry cards;
- capital and shares;
- contacts and activities; and
- certain immovables for building associations.
Those lists may expose multiple dated documents or relationships. They can help reconstruct aspects of an existing company's history. The package still carries the same current-register population restriction, so it is not a full archive of deleted entities.
Nor does the catalogue promise that every list is a complete historical audit trail. “List of articles” or “list of registry cards” should be reported as the fields supplied in the current General data record, not as proof that all prior versions and underlying documents are included.
Registry-card entries provide dated events, not daily snapshots from the past
The Registry cards package is also updated daily. It returns company name and registry code together with entry number, entry date and entry type.
This is event-oriented historical information. Multiple entries can show that register actions occurred on different dates. But the package's current-entity restriction remains. The entries do not establish a complete historical population, and they are not equivalent to preserved copies of the full company record as it appeared on each entry date.
The same caution applies to the Rulings, Persons on registry card, Shareholders and Commercial pledge packages. Dated rows can describe earlier events or relationships for companies still covered. They do not override the catalogue's explicit exclusion of deleted legal entities from the daily open-data files.
Monthly annual-report data is a separate historical series
The is not one of the daily company-master snapshots. Its verified catalogue record states:
- update frequency: monthly;
- indexed file-publication date: 3 July 2026;
- data-coverage end date: 30 June 2026; and
- format: connected CSV files.
The files include general information about submitted reports, sales by activity, sales by geographical region and selected non-consolidated financial indicators. The general-information file includes company name, registry code, submission date, reporting period, status, audit information and possible modifications. Related files connect through report_id.
This is genuine period-linked information, but it is still selective. The catalogue says whole annual reports must be obtained through e‑Business Register services. It also says that reference-period data from reports is not included in the files.
The catalogue conflicts on how many years are available
The prose says that the datasets contain statistics for “the three previous financial years.” The same record then lists separate indicator files for 2019, 2020, 2021, 2022, 2023, 2024 and 2025.
The evidence therefore establishes that links for those seven report years were catalogued in the indexed snapshot. It does not establish why the prose still says three years, whether older files are retained archives, or whether every component file has the same historical scope.
Until is resolved, a researcher should cite the exact year-specific file actually used and avoid the unqualified claim that the open-data series has either exactly three or exactly seven years of uniform coverage.
Statistical queries offer aggregate history
The is a public query surface rather than a bulk entity dataset. The indexed description says it can show aggregate statistics for all Estonian legal entities, including:
- how many legal entities were registered or deleted;
- registered bankruptcy proceedings;
- company employee counts;
- common fields of activity; and
- annual-account filing statistics.
Queries can be made by year or, in the detailed view, by month.
This establishes historical aggregate availability. It does not provide the identity and full history of every entity contributing to a bucket. A statistic saying that a number of entities were deleted in a month cannot be joined to the current Basic data file to identify every deleted company unless another authoritative dataset supplies that relationship.
The portal offers current public queries and possible additional requests
The indexed says that users can search for an entrepreneur by name or registry code and use more detailed search criteria. It also says that written requests can be made to RIK for additional and more specific data groups when needed.
That establishes that access routes exist beyond the bulk downloads. It does not prove that every historical field is freely downloadable, that every additional request will be granted or that contract-based XML/API services are part of the indexed open-data collection.
The Legalcode records expressly exclude contract-only RIK XML/API services unless a no-authentication public endpoint has been verified. The public catalogue and the full commercial or archival service must not be treated as the same product.
Personal-identification coverage changes after 1 November 2024
The catalogue warns that, from 1 November 2024, open-data files no longer include personal identification numbers for data-protection reasons.
That is another historical boundary. A current entity file may still contain names, roles, start and end dates or other relationship information, but users should not expect the public personal identifier previously used for deterministic matching.
Historical joins across that boundary need care. A missing personal code after the policy change is a source-coverage rule, not proof that the person or relationship was absent.
File dates must remain separate from event dates
Several different dates appear in the catalogue:
- source-file publication date;
- retrieval or ingest date;
- dataset coverage end date;
- company first-entry date;
- registry-card entry date;
- ruling date and entry-into-force date;
- officeholder or shareholder start and end dates; and
- annual-report period and submission date.
They are not interchangeable. A daily file dated today can contain a company first registered decades ago. A monthly annual-report package can be published after its coverage end date. A registry entry date is not the date on which the current ZIP was generated.
For reproducible use, each imported row should retain the dataset name, file URL, observed file version, retrieval timestamp and relevant business-event date.
Current Legalcode freshness limitation
On 13 September 2026, the production MCP still reported 5 August 2026 as the publication/file date for the daily Basic and General records. Read-only checks of the exact official ZIP endpoints showed that both files had been modified on 13 September 2026.
The indexed catalogue metadata is therefore more than one month behind the official daily source. tracks the stale refresh and requests a cadence-aware freshness check across all nine daily packages.
Until corrected, the MCP record proves what the indexed catalogue snapshot described; it does not prove that its stored file date is the latest version available from RIK.
An operational history model
A system using these files should keep four evidence layers separate:
If point-in-time reconstruction matters, archive each permitted source snapshot with its checksum or ETag, keep event dates separate and document gaps. Do not backfill a missing historical state by assuming that today's value was always true.
Authority, currentness and MCP audit
The e‑Business Register catalogue is the official description of the public downloads and query surfaces. The production Legalcode MCP indexed 16 catalogue/reference records: nine daily Business Register packages, one monthly annual-report series and six public query or presentation surfaces. It does not index each company row as an individual Legalcode record.
Five catalogue records were fetched through exact MCP source references and their complete authenticated Markdown downloads were checked. The Registry cards record was also verified through the exact download URL returned by Search. All production bypass calls charged zero.
Three defects were confirmed during this question:
- All 16 EE_RIK records are labelled Estonian, although their searchable titles and normalized text are English. The Estonian update/history query returned zero while the English equivalent returned all 16 records. This is .
- The daily catalogue records are stale at 5 August 2026 while the official files were modified on 13 September 2026. This is .
- The annual-report record conflicts between “three previous financial years” and year-specific links covering 2019–2025. This is .
The earlier inventory defect also remains: the repeated source description says there are eight daily Business Register datasets, while nine are indexed. That conflict is tracked in .
Conclusion
The e‑äriregister catalogue describes daily current-state bulk files and a monthly, selective annual-report series. The daily files include only legal entities currently registered, in liquidation or in bankruptcy. They may contain dated events and documents for those entities, but they do not establish a complete historical register and exclude deleted entities from the stated population.
For history, the catalogue offers partial ingredients: dated registry events, reports and relationships for current entities; period-specific annual-report files; and year/month aggregate statistics. None should be misrepresented as a complete point-in-time company database. Reliable historical work requires exact file-version preservation, event-level dates, explicit coverage boundaries and separate evidence for deleted entities.
Sources and audit issues
This article provides general legal information, not advice for a specific matter. Laws, facts and procedural deadlines may change the result. Follow the linked primary sources and obtain local advice before acting.