This project explores the possibility of risk selection in Medicare managed care plans through the supplemental benefits packages offered. It seeks to understand if consumers are differentially attracted to different benefits packages in the basis of their personal health status. If there is strong evidence of risk selection through supplemental benefits packages, then plans are doing a poor job of pooling risk, and relatively sick enrollees will not find plans attractive to them. If there is little evidence of risk selection then proposals to limit the supplemental benefits packages offered by plans will reduce consumer choice, and thus reduce welfare. Discrete choice models of consumer demand will be used to determine the existence of risk selection. Consumer choice of plan will be modeled as a function of individual characteristics, such as health status; plan characteristics, such as benefits packages; and market characteristics such as Medicare reimbursement rate. The models will determine which, if any, benefits packages lead to risk selection. They will also determine if risk selection exists at different Medicare reimbursement rates. Finally, a welfare analysis of policy proposals aimed at reducing risk selection will be conducted.