The basic data format of PDT 2.0 is PML ("Prague Markup Language"), which is based on XML. Formerly two other formats were used to analyze and save PDT data. The format FS ("Feature Structure") was developed for the programme Netgraph (or rather for its predecesssor, i.e. programme Graph). The basic format of PDT 1.0 was CSTS ("Czech Sentence Tree Structure"), based on SGML. Nowadays, this format is used only as a work format for older NLP tools (e.g. parsery and tagery).

For details on individual formats see Prague Dependency Treebank 2.0, CDROM, doc/pdt-guide/ and doc/data-formats/.

For more on the programme Netgraph see Prague Dependency Treebank 2.0, CDROM, doc/tools/netgraph/.