Elderly persons with functional limitations face long term care (LTC) policy alternatives that are largely determined at the state level. States vary greatly in their efforts to transform LTC systems that rely almost exclusively on institutional services to those that provide services in alternative settings including home and community-based services (HCBS). The proposed study has two broad objectives: (1) to investigate how trajectories of LTC use and well being among elderly persons vary, as a function of state level policies and across cohorts, and (2) to obtain in-depth understanding of state LTC policy contexts. The first objective will be achieved in Part I of the proposed study, which centers on multilevel modeling of state-level LTC policies and individual-level trajectories of LTC use and well being. The analysis will use longitudinal data from the Health and Retirement Study (HRS,1993, 1995, 1998, 2000, and 2002) and National Long-Term Care Survey (NLTCS, 1989, 1994, and1999), which will be merged with state-level LTC policy data, using the state identifiers available in the HRS and NLTCS data sets. Since LTC policies and the characteristics of elderly cohorts needing care change over time, Part I also involves the compilation of state-level LTC policy data that cover the survey years of HRS and NLTCS as well as multiple cohort analysis. The second objective will be achieved in Part II, which consists of case studies of 2 states (New York and Illinois). Both states are similar in that they have a major metropolitan area with diverse populations as well as large upstate or downstate rural areas, but differ markedly in terms of political background, LTC systems, HCBS expenditures, and programs/services offered. In-depth state policy contextual information will be obtained through site visits (interviews with state officials, LTC provider associations, advocacy group representatives and other stakeholders) and from public documents, websites, and electronic data files. In addition, trajectories of LTC utilization and well being among HRS and NLTCS respondents of the 2 states will be described and compared. Part I and Part II will proceed simultaneously to inform each other throughout the four-year study period.