Six-month and lifetime rates of DSM-III major depressive disorder (MDD) and characteristics of the disorder were compared in mothers of children with disabilities (chronic stress sample, n = 310) and in a geographically based probability sample (controls, n = 357). The presence of DSM-III MDD was ascertained by the National Institute of Mental Health Diagnostic Interview Schedule. Although mothers in the chronic stress sample had significantly more depressive symptoms, rates of MDD were not significantly different in the two samples. The women with MDD in the chronic stress sample reported a lower age of onset and more episodes in lifetime than diagnostically comparable controls, but the two samples did not differ in symptomatology of worst episode. The data do not support an etiologic role for chronic stress in MDD, nor do they support the hypothesis that chronic stress is associated with a unique symptom profile or more severe episodes. They suggest a role for chronic stress in precipitating episodes, although the evidence on this point must be interpreted with caution.