The literature on the reduction of aggression in people with learning difficulties is reviewed. The methods of reducing aggression identified were predominantly evaluated in staffed settings and with target behaviours more frequent than once a day. This contrasts with the findings of a recent survey that found the majority of aggressive individuals to live in the community and show aggression less than once a day. It is argued that the reasons for this disparity are as follows. That it is easier to do a controlled study in staffed settings with high‐frequency aggression. That it is more difficult to do an analysis of the target behaviour if it is low frequency or if the subject is in the community; and that the methods that have been traditionally used to reduce aggression may not be effective with all subjects in all settings. It is suggested that it may be necessary to develop and evaluate a new technology for reducing aggressio