This dissertation examines the quality, cost, and enrollment effects of managed care in the Medicare program between 1999 and 2004. Administrative enrollment, spending and inpatient hospitalization data are analyzed. During the study period, Medicare managed care enrollees are positively selected relative to those who remain in Fee-for-Service, with fewer total comorbid conditions and lower rates of several chronic conditions. Multiple econometric strategies are used to address this selection in the remaining chapters. I find that Medicare managed care plans and traditional (Fee-for-Service) Medicare provide similar levels of quality and access to care on many measurements. Managed care plans outperform Fee-for-Service in reducing rates of potentially preventable hospitalizations that can be avoided with timely antibiotic use. Using an instrumental variables approach, I show that a one percentage point increase in Medicare managed care penetration increases total county Medicare spending by 1.1 percent, $2.6 billion in 2004 dollars. Findings have important implications for policymakers considering further Medicare reforms as well as beneficiaries and their families choosing coverage options.