The present article is a corpus-driven investigation into deontic modality in the Lithuanian translations of EU legislation. It builds on the data of the EUR-Lex2/2016 parallel corpus. The article discusses the pivotal points of deontic modality and its realisation in English and Lithuanian. It then presents the corpus data and the most frequent lexical bundles through which deontic modality is expressed in the English subcorpus. The discussion section analyses how deontic lexical bundles in the source text were translated into Lithuanian.