This chapter highlights the relevance of text categorisation for research in legal translation by focusing on institutional translation settings, namely: the European Union (EU), the United Nations (UN) and the World Trade Organization (WTO), and their corresponding adjudicative bodies.1 After briefly reviewing recurrent issues and models of legal text classification (section 2), a multidimensional approach is applied to the multilingual text production of the three representative institutional translation settings during three years over the span of a decade (2005, 2010 and 2015), as part of the project "Legal Translation in International Institutional Settings: Scope, Strategies and Quality Markers" (LETRINT) (section 3). The resulting subdivisions are integrated into a categorisation matrix and discussed as a way of illustrating the relative nature and implications of text classifications. The fine-grained description of corpus design and representativeness, technical aspects of corpus compilation and full taxonomies of genres are not addressed in this chapter.