Background: Little is known about the independent associations of reward and stress within specific roles with multiple measures of mental health in an ethnically diverse community sample of midlife women. The objective of this study is to examine if (1) role reward (within each role and across roles) contributes directly to mental health and buffers the negative impact of role stress and (2) associations among role occupancy, role stress, and role reward and mental health vary by race/ethnicity. Methods: With separate logistic regression analysis, we investigated cross-sectional relationships between role stress and role reward with presence/absence of high depressive symptoms (Center for Epidemiologic Studies Depression Scale [CES-D ‡ 16]), anxiety symptoms (feeling tense or nervous, irritable or grouchy, fearful for no reason, and heart pounding or racing total score ‡ 4), or low social functioning (bottom 25th percentile of the Short-Form-36 [SF-36] social functioning subscale) in 2549 women participating in the third visit of the Study of Women's Health Across the Nation (SWAN), a longitudinal population-based study of menopause. Results: High reward across roles attenuated the negative impact of role stress on social functioning but not on anxiety or depression. High reward marriage buffered the impact of marital stress on depression, and high reward mothering buffered the effect of maternal stress on depression and social functioning. Compared to Caucasians, Hispanics and Chinese with high stress across roles had better social functioning, and African American mothers had lower odds of high depressive symptoms. Conclusions: Role reward buffers the negative impact of stress on social functioning and depression, but not on anxiety. Minorities may respond to role stress by seeking social support.