鈥� To address the validity of subtype distinctions within a large family study of major depression, probands (N=133) were classified into several non-mutually exclusive subcategories, including endogenous (n = 89), melancholic (n=61), autonomous (n = 50), and delusional (n=21). Agecorrected lifetime rates of depression and subtypes among first-degree relatives were then compared by the proband's depression subtype. Rates of major depression were highest for the relatives of probands with the autonomous and delusional subtypes, and while lower for the relatives of endogenous and melancholic probands, these rates were still higher than for the relatives of the remaining depressed probands or the relatives of normal controls. The depressed relatives of depressed probands with the endogenous, melancholic, autonomous, or delusional subtypes were more likely to have one of these subtypes than the depressed relatives of either the remaining depressed probands or the normal controls.