We examine the relationship between ancestral age at death and affective disorder. A cohort of 204 socioeconomically favored men was selected for mental and physical health 50 years ago and followed up until age 69 +/- 1 (+/- SD) years in an interdisciplinary study. During follow-up, 49 of these men were identified at some point as being psychosocially impaired. Of these 49 men, 25 were identified as having probable affective disorder based on family history of affective illness, clinical signs, subjective symptoms, clinical diagnosis, and choice of pharmacological treatment. Objective signs and symptoms distinguished these 25 men from the remaining 24 men, whose adjustment over the 50 years was equally psychosocially impaired, but who were never noted to be significantly depressed. The mean age at death of the maternal grandfathers for the 25 depressed men was 60.4 years, significantly younger than the mean age at death of maternal grandfathers for either the total sample (70.1 years) or the 25 psychosocially impaired but not depressed men (68.8 years). The mean age at death of the depressed men's other five first-degree ancestors was not significantly different from the age at death of the ancestors of the rest of the sample. If depressed alcohol abusers were excluded, the mean age at death of maternal grandfathers of alcohol abusers did not differ from that of the controls' maternal grandfathers. The evidence for possible X-chromosome linkage in male psychobiological vulnerability to affective spectrum disorder is discussed.