The effect of stressful events on depression has been amply demonstrated, but the opposite relation is also important. I examined event occurrence over 1 year in 14 women with unipolar depression who were compared with demographically matched groups of women with bipolar disorder (n = 11), chronic medical illness (n = 13), or no illness or disorder (n = 22). Interview assessments of life events, severity, and independence of occurrence confirmed the hypothesis that unipolar women were exposed to more stress than the normal women, had significantly more interpersonal event stress than all others, and tended to have more dependent events than the others. The implication is that unipolar women by their symptoms, behaviors, characteristics, and social context generate stressful conditions, primarily interpersonal, that have the potential for contributing to the cycle of symptoms and stress that create chronic or intermittent depression.