This thesis presents a definition of eurolect and a corpus-based analysis of Czech eurolect. The theoretical part of the thesis focuses mainly on the distinction between existing concepts of the specific language of translations originating in EU institutions and on the definition of its main characteristics. As the process of translation in EU institutions is unprecedented, this thesis provides its detailed description. The analysis of Czech eurolect and its main features was conducted on several sub-corpora comprising of EU translations and reference corpora comprising of Czech journalistic texts from before and after 2004.