The variations of cancer mortality rates from place to place reflect both underlying differences in cancer prevalence, as well as differences in diagnosis and treatment that affect the risk of death. The latter are particularly important in explaining the less favorable outcomes for disadvantaged populations and racial and ethnic minorities. The objectives of this study are to explore geographic variation in late-stage cancer diagnosis within Illinois and to examine the role of access to healthcare in explaining such variation. Using geographic information systems (GIS) and innovative spatial analysis methods, this research will (1) describe and analyze spatial clustering of late-stage cancer, both for the population as a whole and by race/ethnicity; (2) model geographic variation in spatial access to healthcare services with an emphasis on screening services for early detection; and (3) examine the effects of spatial access to healthcare along with various non-spatial factors in influencing geographic patterns of late-stage cancer. Specifically, we will focus on three of the four leading types of cancer in Illinois: breast, colorectal and prostate cancers, all of which have relatively effective means of early detection. All analyses will be conducted for the population as a whole and for African-Americans and Latinos separately to shed light on race and ethnic disparities. The research will provide an informed basis for targeting health education and cancer prevention services to the areas where they are most needed.