With the use of family study methods and assessments by "blinded" raters, we tested hypotheses about patterns of familial association between DSM-III attention deficit disorder (ADD) and affective disorders (AFFs) among first-degree relatives of clinically referred children and adolescents with ADD (73 probands, 264 relatives) and normal controls (26 probands, 92 relatives). Among the 73 ADD probands, 24 (33%) met criteria for AFFs (major depression, n = 15 [21%]; bipolar disorder, n = 8 [11%]; and dysthymia, n = 1 [1%]). After stratification of the ADD sample into those with AFFs (ADD + AFF) and those without AFF (ADD), familial risk analyses revealed the following: (1) the relatives of each ADD proband subgroup were at significantly greater risk for ADD than were relatives of normal controls; (2) the age-corrected morbidity risk for ADD was not significantly different between relatives of ADD and ADD + AFF (27% vs 22%); however, these two risks were significantly greater than the risk to relatives of normal controls (5%); (3) the risk for any AFF (bipolar disorder, major depressive disorder, or dysthymia) was not significantly different between relatives of ADD probands and ADD + AFF probands (28% and 25%), but these two risks were significantly greater than the risk to relatives of normal controls (4%); (4) ADD and AFFs did not cosegregate within families; and (5) there was no evidence for nonrandom mating. These findings are consistent with the hypothesis that ADD and AFFs may share common familial vulnerabilities.