FISCAL SOLUTIONS...

Czech Republic: Financial Administration publishes CSV format for bulk import of EET 2.0 evidence units

Add to Favorites Leave a Comment
Summary

The Czech Financial Administration has published the CSV structure for bulk import of EET 2.0 evidence units into DIS+. The 16-column file lets businesses with many premises, vending machines, websites or vehicles register them in one step before requesting cash register certificates.

Content

The Czech Financial Administration has published a new document titled "Import evidenčních jednotek EET – popis struktury CSV souboru" (Import of EET Evidence Units – Description of the CSV File Structure) on eet.gov.cz. The four-page document, issued by the General Financial Directorate and dated 3 November 2026, is listed both among the documents for developers and among the information materials for EET 2.0.

The document follows the launch of EET 2.0 functions in the tax information box DIS+ on the MOJE daně portal, announced by the tax administration on 2 November 2026. Since then, entrepreneurs can register for sales recording, set up evidence units and generate cash register certificates in preparation for regular operation from 1 January 2027.

Legal background

Under Section 15 of Act No. 180/2026 Coll., on the Recording of Sales (ZoET), a taxpayer must report data on all its evidence units through DIS+ before obtaining a certificate, if recorded sales are made through at least one of these units. The document states that the tax administrator allows this data to be reported by bulk import. In DIS+, under the "Evidence tržeb" tab, the "Správa evidenčních jednotek" (Management of Evidence Units) page offers manual creation of a unit or a "Načíst ze souboru" (Load from file) button for importing a prepared CSV file.

File format rules

The document sets the following technical rules for the import file:

  • The file is in CSV format with a semicolon (;) as separator, one line per evidence unit and a fixed order of fields.
  • Accepted encodings are UTF-8, ISO 8859-2 and Windows-1250.
  • Text values must not contain the characters ";" or "|", or characters that cannot be stored in ISO 8859-2.
  • A file may contain at most 500 lines, with up to 5,000 characters per line, and must not exceed 5 MB.
  • A line must have at least 7 and at most 16 columns. Missing trailing columns are treated as empty, but empty columns in the middle of a line must be kept.
  • The file must not contain a header row. According to the document, support for a header row is being prepared, and the description will be updated once it is deployed.

Structure of the record

Each line describes one evidence unit in 16 columns. The mandatory fields are the type of evidence unit, the predominant business activity, and two flags stating whether recorded sales are made in the unit and whether the taxpayer accepts sales in indirect representation. The type of unit is a single letter: S (fixed premises), M (mobile premises), A (vending machine), I (website or application), D (vehicle) or P (taxpayer). The business activity is entered as a CZ-NACE code from the current "Činnosti" codelist.

The location depends on the type of unit. For fixed premises, vending machines and the "taxpayer" type, an address can be given. The preferred option is the RÚIAN address place code. If this code is filled in, the separate address fields (street, building number, orientation number, municipality, part of municipality and postal code) are ignored. If no RÚIAN code is given, the building number, municipality and postal code are mandatory for fixed premises and the "taxpayer" type. For a vending machine, they are mandatory only if other address details are entered.

A free-text location field is mandatory for mobile premises (a description of the location), websites and applications (a URL or other unique identifier) and vehicles (a registration plate or other identifier). It is also mandatory for a vending machine without an address. For vehicles, the vehicle type is mandatory. If "other" is selected, a short specification must be added.

In our view

The import is mainly relevant for businesses with a large number of evidence units, such as retail chains, operators of vending machine networks, transport companies and ecommerce businesses with several sales channels. In our view, the most practical approach is to generate the file from the client's existing master data, for example a store or asset register, rather than prepare it manually. This is our recommendation, not a requirement of the document.

Several points need attention. The 500-line limit means that larger networks will have to split their data into several files. The absence of a header row and the fixed column order leave little room for error, so a column shift will lead to an incorrect import. Using RÚIAN codes reduces the risk of address mismatches, because the code takes precedence over the text address fields. The document is a technical description of the file and does not state how validation errors are reported. It also does not describe how existing units are updated or deleted through the import. These questions remain open and should be clarified with the tax administration if they matter for the client's rollout.

What should businesses and vendors do?

Businesses with many evidence units, and the vendors or consultants who support them, should prepare their data now:

  • Make an inventory of all evidence units and decide for each one the correct type, CZ-NACE activity code and whether recorded sales are made through it.
  • Obtain RÚIAN address place codes for fixed premises and vending machines where possible.
  • Prepare identifiers for websites, applications and vehicles, such as URLs and registration plates.
  • Build the CSV export in the required encoding, without a header row and with the fixed column order, and split it into files of up to 500 lines.
  • Test the import in DIS+ with a small file first, then check the created units before requesting cash register certificates.
  • Keep an internal mapping between the client's own location IDs and the evidence units, as the evidence unit is used in the EET 2.0 data message.

This article is based on the document "Import evidenčních jednotek EET – popis struktury CSV souboru" and related information published on eet.gov.cz and the MOJE daně support portal - Source Source

 

 

Aleksandra Vukić


Business Analyst
Linkedin profile

Comments

Questions and comments (0)

There are no comments on this news yet.

The latest 3 updates: