Groups of home and school situational hyperactive primary schoolboys identified from the community were compared with pervasive hyperactive and non-hyperactive controls on a wide range of measures. The hyperactive groups tended to persist in the same category over a half-year period. Both situational hyperactive groups had lower measured activity levels than the pervasive hyperactive group and only the latter differed from non-hyperactive controls. Home hyperactivity was characterized by poor family relationships and was not distinguishable from non-hyperactive home-antisocial controls. School hyperactive boys had specific correlates of low intelligence, motor clumsiness, poor reading and academic abilities. Pervasive hyperactive subjects differed from both situational groups in showing a higher percentage of delayed language development. While home hyperactivity has dubious identity, the distinct pattern of external correlates in school and pervasive hyperactivity speak for the need to regard these as separate entities.