1. Experience from scale validation studies and from controlled drug trials provided clues for optimalisation of studies in aggressive patients. 2. Definitions of target behaviour, selection of patients and measurement of aggression are reviewed and recommendations are presented.