This paper examines psychiatric diagnosis-making within biomedicine, as studies in this field, especially empirical ones, have attracted little attention throughout the scientific world. The widely accepted assertion, that diagnosis is the result of clinical judgement, will be discussed. Through analysis of a case study conducted in a Barcelona hospital, it will be suggested that psychiatric diagnosis is not only the result of clinical assessment, but also of psychotropic drugs translation. Drawing on actor-network theory and the notion of boundary objects established by Star and Griesemer (1989), it is therefore proposed that psychotropic drugs are boundary objects which act as central mediators in knowledge management, setting up the semioticmaterial assemblage of diagnosis.