The paper uses data from nursing-home cost reports to analyze the effectiveness of different approaches to nursing-home reimbursement. Our research has produced considerable evidence on the effect of states' efforts to reduce the rate of increase in nursing-home costs. First, homes in states with flat-rate reimbursement systems were found to have lower rates of increase than homes in other states, while there were no consistent differences between the results of prospective and retrospective systems. Second, efficiency incentives, inflation-projection methods, and the level of ceilings on rates appear to be very important, regardless of the general reimbursement method. For example, prospective systems with weak efficiency incentives, generous inflation adjustments, and high percentile ceilings have lesser cost-containment effects than prospective systems with stringent inflation allowances and low percentile ceilings. There is also evidence that the inherent weakness of the cost-containment incentives in retrospective systems can be offset by low percentile ceilings and efficiency bonuses.