Demographic, psychiatric, social, cognitive, and life stress variables were used to determine the etiology of depression in childbearing (CB; n = 182) and nonchildbearing (NCB; n = 179) women. Hormonal variables in postpartum depression were also evaluated. In the CB group predictors of depression diagnosis were previous depression, depression during pregnancy, and a Vulnerability (V) x Life Stress (LS) interaction; predictors of depressive symptomatology were previous depression, depressive symptoms during pregnancy, life events, and V x LS. Only estradiol was associated with postpartum depression diagnosis. In the NCB group V X LS was the only predictor of depression diagnosis; depressive symptoms during pregnancy and life events were predictors of depressive symptomatology. Previous findings about depression vulnerability were replicated. The significant V x LS interactions support the vulnerability-stress model of postpartum depression.