In order to contain the escalation of health care expenditure, the payment system of the National Health Insurance (NHI) has been gradually transformed from a fee-for-service (FFS) basis to prospective payments (case-payment). The case-payment system was first experimented with three items in 1995 as a demonstration, progressively increasing to 54 items in 2004. It is well documented in the health economics literature that health insurance creates behavioral change for providers. This study will evaluate the impacts of the case payment system and, furthermore, the effects of resource utilization patterns in hospitals. In order to study hospital practices, claims of hospital discharges were extracted from NHI’s inpatient expenditures and detailed orders files from 1997 to 2002. The case payment items were classified into two groups,1997 and 1999 groups, according to the year of implementation. FFS cases were the reference group. A comparison of the three groups was conducted using the Ashby’s performance model. The production of a discharge is decomposed into three components: the number of days, the service content of each day, and the inputs required to produce each unit of service. Changes in these measurements were adjusted for deflation and calculated. Major findings are as follows: 1.Effects on health expenditure: For those diseases paid by case payment, average medical treatment expenditure showed a decreasing trend. For those diseases paid by case payment in 1997, the rate of reduction for average medical treatment expenditure was comparable to the rate of reduction for average length of stay. While, for those diseases paid by case payment in 1999, the reduction rate for average medical treatment expenses was markedly smaller than the reduction rate for average length of stay. 2.Effects on length of stay: After implementation of the case payment system, the average length of stay was shortened. Concerning the diseases paid by FFS, the length of stay tended to increase slightly. 3.Effects on volume of medical procedures：After implementation of the case payment system, the intensity of medical treatment services showed no decreasing trend. 4.Effects on different hospitals: a) By ownership category: Non-profit proprietary hospitals appear to be more efficient than public and private hospitals. As for FFS cases, private hospitals increase revenues by increasing the average length of stay. Non-profit proprietary hospitals increase revenues through a similar strategy used for case payment cases. b)By accreditation category：For those diseases paid by case payment , average total medical claims for regional hospitals were less than for medical centers, medical claims for district hospitals were less than for regional hospitals. For those diseases paid by FFS , the rate of increasing for average medical expenditure for regional hospitals and medical centers were less than the range of increasing for average medical expenditure for district hospitals. In addition to providing a comparative reference on the effects of policy implementation, the results of this research indicate two possible directions for future studies: 1.By combining the outpatient service data, an attempt can be made to determine whether the decrease in expenditures after the implementation of a case payment system is due to cost shifting to outpatient services. 2.The results of 2002 indicate that medical expenditures increase in both. It is therefore interesting to investigate whether there exists relationships between this phenomenon and the implementation of the global budgeting system in the hospital sector.