Background: National and regional data indicate that breast cancer early detection is low in Kentucky, especially rural regions, perhaps because access to mammography services can be problematic. Objective: This study examined the distance between residences of women diagnosed with breast cancer and the nearest mammography facility, as a risk factor for advanced stage diagnosis in rural populations. Methods: 1999-2003 Kentucky Cancer Registry data were used for this study. A total of 12,322 women, aged 40 and older at diagnosis, with no previous history of cancer, and with known cancer stage were included. Travel distance was obtained using a geographic information system (GIS). Hierarchical logistic regression models were used to analyze the relationship between travel distance and advanced stage diagnosis. Results: Advanced diagnoses had longer average travel distances than early stage diagnoses (P less than 0.01). After adjusting for age, race, insurance, and education at census tract level, the odds of advanced diagnosis were significantly greater for women residing over 15 miles from a facility, compared to those living within 5 miles (adjusted OR = 1.50, 95% CI = 1.25-1.80). Conclusion: Although socioeconomic status, race, and age may help explain advanced diagnoses, longer travel distance also adversely affects early detection for rural populations. Accurate measurement of spatial accessibility indicators, such as travel distance, facilitates identification of at-risk groups so that interventions can be developed to reduce this disease.