Rechtspraak

Rechtspraak publishes the rulings of the Dutch judiciary. This is the only source in BesluitBron that is not a decision-making register. The other six set out what an administrative body has decided; this one sets out what a judge has ruled on the matter.

Two services, not one

What is known as ‘Rechtspraak’ consists, in practice, of two services with different characteristics. Determined on 15 and 16 August 2026 in approximately 300 measurements.

flowchart TD
    BB["BesluitBron"] --> R["data.rechtspraak.nl<br/>register and<br/>price lists"]
    BB --> S["rulings.rechtspraak.nl<br/>search within the text"]
    R --- N["3,743,498 registrations<br/>no text searches"]
    S --- P["951,718 published<br/>rulings<br/>text, facets,<br/>phrases"]

The two services agree, down to the last unit, on the size of the published section. The register lists 951,718 ECLIs associated with a document, and a search without a search term returns 951,718 results. That is the strongest evidence that both figures are exact. It is also the reason why an ECLI found on one service can be retrieved on the other.

What’s inside

• 3,743,498 ECLIs, comprising 3,668,096 rulings and 75,402 opinions of the Advocate General, plus 45,920 records of foreign decisions;
• 951,718 of these contain a published text, i.e. 25.4 per cent.

The most significant limitation

Three-quarters of the registered rulings have not been published, and this is not a backlog. Publication is a decision taken by the court. The missing three-quarters will therefore not be made available at a later date.

It follows that the only correct way to phrase a negative result is: “nothing was found on this matter in the published rulings”, and not “no ruling was made on this matter”. See Zoekprecisie.

Furthermore, the coverage has increased significantly over time. In 1980, 1.7 per cent of entries contained text; in 1995, 6.7 per cent; and in 2000, 34.2 per cent. Consequently, a search for older legislation systematically misses more results than a search for more recent legislation.

Other restrictions

• The search service is not documented and is not supported. BesluitBron uses it because there is no other way to search the text, and treats any changes to it as a risk associated with this connector;
• The register does not support any form of text search. A query based on a word in the ruling can only be answered by the search service;
• Neither service offers compression or HTTP/2. As a result, retrieving a large number of rulings in quick succession is slow.

Address and tools

Address: https://besluitbron.nl/mcp/rechtspraak/v1.

• search_rulings: search the text of published rulings, using filters and phrases;
• find_rulings: search the register by criteria such as organisation, legal area and period, without text searches;
• get_ruling: a single ruling with its metadata, on ECLI;
• get_ruling_text: the full text of a ruling;
• related_rulingss: related rulings, via references;
• recent_rulings: recent publications;
• list_courts: the authorities, taken from the register’s own list of values;
• list_legal_areas: areas of law;
• list_procedure_types: types of proceedings;
• data_freshness: the date up to which the publication is up to date;
• list_platform_calls: the calls to the platform in this session.

In addition, list_capabilities.

Which tool is used for which service

search_rulings uses the search service and returns 951,718 rulings with a published text. find_rulings uses the register and returns all 3,743,498 entries, but cannot search within the text.

The distinction is not technical but substantive. The first tool answers the question ‘where does this word appear in the rulings?’. The second answers the question ‘which rulings by this body were recorded during this period?’.

Search by time period

search_rulings Both find_rulings and date_to contain the dates date_from and . These are the start date and end date of the period, both in the format jjjj-mm-dd. The dates refer to the day of the ruling, not the day of publication. These two dates are often weeks apart.

The two dates are independent of each other. A start date alone means: from that day until now. An end date alone means: everything up to and including that day. Both dates are included in the calculation.

This is also correct from version 1.3.39 onwards. Up to and including version 1.3.38, search_rulings filled in the missing side with the date from the other side. As a result, a search using only date_from would search within a single day rather than over a period. The result was usually zero rulings, with no indication that anything was wrong. Anyone searching on an older version should therefore always specify both dates. find_rulings did not have this problem.

Since version 1.3.39, the response from search_rulings has included the field date_range_sent. This field contains the time period that BesluitBron actually sent to the search service, in that service’s notation (dd-mm-jjjj). The field is empty if no date filter has been specified. This means that, in the case of an empty response, it is possible to see whether the filter is the cause.

Where to find this platform

These addresses belong to the organisation behind this source and not to BesluitBron. Anyone who finds an error in the data or in this source’s API should report it there:

• Report an error: the dataset page on data.overheid.nl ;
• Reading the API: Open Data from Rechtspraak ;
• Search for yourself on the source’s website: uitspraken.rechtspraak.nl .

Rechtspraak does not provide support for using the web service and does not have an issue tracker. However, the dataset page does state who to contact with any queries. The documentation page rejects programmes with an error code and simply opens in a browser.

These addresses were checked on 18 August 2026. No email address has been included deliberately: where a party publishes only an address, BesluitBron refers to the page on which that address appears.

What this source is suitable for

To find out how the court has ruled on an administrative decision. And to find comparable cases when preparing a decision or a defence.

What it is not for

To determine the court’s general position. The reason is set out above: three-quarters of registered rulings are not published, and for older case law, that proportion is even higher.

Current situation, as of August 2026

There is a discrepancy between these two timekeepers, and that difference is the most important thing a user needs to know.

The feed is being updated. On 16 August 2026, the most recent published ruling had been amended the previous day, and 386 entries had been amended in 24 hours.

The content is months behind. Rulings from a single Wednesday, counted by published text: 92 after three days, 148 after ten days, 268 after seventeen days, 374 after a month and 543 after three months. The curve was still rising after three months. Any response regarding last week’s rulings is therefore systematically incomplete. This is a characteristic of the publication itself and not of this connector.