Where and how one lives is associated with cancer survival. This study was designed to assess geographical region of residence, race/ethnicity, and clinical and socioeconomic factors as predictors of survival in a population based cohort of women with breast cancer followed for up to 12 years. In a cohort of 218,879 breast cancer patients >20 years of age at diagnosis, registered in the database of the US National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program between 1990 and 2001, we analyzed the association of breast cancer-specific survival with SEER region; age; stage; histology; hormone receptor status; race/ethnicity; and census data on educational attainment, income, employment, and insurance coverage. We compared Kaplan-Meier survival curves by region and race/ethnicity. We used Cox proportional hazards regression models to assess the association of mortality with region, race/ethnicity, and the other variables. Women who lived in Detroit had significantly higher mortality than those living in most other SEER regions. In most regions, black women had the poorest survival. The association of mortality with race did not differ significantly across regions, but it was significantly stronger among women 50-64 years of age than among women 65 and older. The SEER data document the association of breast cancer mortality with region, race, and socioeconomic status. Black race was a strong predictor of mortality in each region even after controlling for socioeconomic factors. The diminishing effect of race with age, which may only partially be explained by insurance in those over 65, suggests a need for research on the role of other factors, such as comorbid conditions or access to care, in breast cancer mortality.