The principle aim is to investigate health disparities in the older population by race, ethnicity, and socioeconomic groups in the United States. The analyses develop and apply random-effect and fixed-effect frailty models for the investigation of mortality differentials and mortality selection processes in mid-and late-life at adult and old ages. These methods will be applied to data from the “Berkeley Mortality Database”, the Social Security Administration, the Health and Retirement Study (HRS), the National Health Interview Survey (NHIS). The research will yield insights into the differential increase in the mortality and morbidity trajectories for various sub-groups in the U.S. population. In particular, the research will investigate the selection dynamics that underlie the observed patterns of mortality by sex, race, ethnicity and socioeconomic groups and assess how the effects of various individual-level characteristics change with increasing age. In addition, the research proposes to use a different approach – the time-to-death approach – to measure health and morbidity differentials by race and ethnicity.