<p> Introduction: Analyzing trends in breast cancer mortality can ensure a precise characterization of changes over time and can be important in public health decision making. Most reported trends are limited to incidence and mortality rates among Whites and Blacks, without categorization regarding tumor clinical characteristics. This study analyzed breast cancer mortality trends among different race-ethnic groups using various approaches such as partitioning rates by factors associated at the time of diagnosis; taking into consideration age, cohort and period effects; and by evaluating geographical variations.</p><p> Methods: Incidence and mortality data from 1990 to 2012 of female invasive breast cancer among women aged 18-84 years in United States (U.S.) was provided by the National Cancer Institute. The following analyses were conducted: (1) calculation of incidence based mortality (IBM) rates by estrogen receptor (ER) status according to race-ethnicity; (2) examination of temporal trends using age-period-cohort (APC) analysis on incidence and mortality rates; and, (3) spatiotemporal analysis of the county level age-standardized breast cancer mortality rates to identify significant geographical areas with higher risk. </p><p> Results: IBM rates for ER+ tumors increased while those of ER- tumors decreased among all race-ethnic groups. APC analysis showed that race-ethnic disparities were largely among the ER- tumors and temporal trends of the ER+ tumors were similar across the race-ethnic groups, with identical effects across the various birth cohorts. Geographical variation in the breast cancer county-level mortality rate was mostly explained by age-standardization and county level risk factors, although the effect of these factors was greater in rural areas of western U.S.</p><p> Conclusion: Temporal trends in the IBM rates were more reflective of the recent changes in the incidence trends of female invasive breast cancer. Trends of ER+ tumors were similar across all race-ethnic groups suggesting a common risk factor for the persistent increase in the incidence and mortality of these tumors. Spatial analysis shows that the higher mortality risk in certain rural counties of western U.S. might be due to poor survival than an elevated incidence and the need for better health care access in these medically underserved areas. These results might explain the observed ethnic and geographic variations in breast cancer mortality, and in turn, could support a stronger theoretical basis for public health policy.</p>