Using criterion-based structured interview techniques and blind assessment, the authors reported high rates of DSM-III childhood attention deficit disorder (ADD) in nonpatient first-degree relatives (parents and siblings) of 6- to 17-year-old clinically referred probands with ADD compared with rates found in relatives of normal comparison children of the same age. To further examine the validity of the ADD diagnosis in these nonreferred relatives, the authors examined whether the diagnosis was associated with antisocial disorders known to co-occur with ADD. As predicted, they found that relatives with childhood ADD (32% of relatives) were at a significantly higher risk for antisocial disorders (61% vs. 19%, p less than .01) compared with relatives without childhood ADD. In addition, the retrospective diagnosis of ADD among these nonreferred relatives resulted in a pattern of observations that is consistent with the literature on ADD in clinically referred children and adolescents: (1) ADD was more common among males than females; (2) the rate of ADD in a control group was consistent with the known risk of ADD in the general population (5.7%); and (3) 71% of ADD relatives reported levels of symptomatology within the range found in clinically referred children. These findings in a group of unselected and blindly evaluated relatives of ADD children provide indirect support for the validity of the diagnosis of ADD using standardized instruments and operational criteria.