OBJECTIVES To examine, using nationally representative data, the impact of Medicare Part D on medication utilization, out-of-pocket-costs, non-drug medical service utilization, and health outcomes among civilian non-institutionalized Medicare beneficiaries without prescription drug coverage prior to the implementation of Medicare Part D. METHODS The primary data were from the Medical Expenditure Panel Survey (MEPS) panel 10 data and the corresponding 2005 and 2006 pharmacy component data. This study included Medicare beneficiaries between 65 and 79 years old that had only Medicare Fee for Service (FFS) in 2005. Patients between 55 and 63 years old with no drug coverage in 2005 served as control subjects. Difference-in-differences was used to identify the effects of Medicare Part D on health care utilization, expenditures, non-drug medical service utilization, and health outcomes among Medicare beneficiaries. RESULTS The mean amount paid by the near elderly without insurance in 2005 declined 7.8%, while the mean amount paid by Medicare beneficiaries with Medicare FFS only in 2005 decreased 40.6%. The number of prescription refills of the near elderly increased 10.35%, while there was only 4.3% increase among the Medicare beneficiaries. ER visits among the Medicare FFS group declined 2.6%, while the number of ER visits among the near elderly increased 0.94%. There was only small growth rate difference in other outcomes variable. CONCLUSIONS Evidence from the MEPS data set suggests that Medicare Part D reduced out of pocket costs, drug utilization, and ER visits among patients without prior coverage. While cost reduction was expected, the reduction in utilization of drugs was not. Future research should further examine these trends.