Since the implementation of the National Health Insurance (NHI) in Taiwan, the NHI expenditure has grown dramatically. In order to make the most of the precious NHI resources and improve the medical care quality of diabetes mellitus (DM), the Bureau of the NHI (BNHI) has implemented the ¡§pay-by-quality¡¨ (PBQ) system in an attempt to stabilize the financial viability and increase the medical care quality since October, 2001. They incorporated the indicators of medical care quality to evaluate the medical institutions before making a reasonable insurance payment to them. BNHI encouraged hospitals to form the diabetic care network to improve the diabetic care quality and utilize the NHI resources efficiently with the cooperation of medical professionals. A successfully established network is believed to be beneficial to all the concerned parties, including the patients, hospitals, and the NHI. It is noted that the effectiveness of the policy requires further studies. Therefore, this research aimed to examine two subjects with statistical analyses. It attempted to identify the effects of the PBQ system on the outpatient medical resources expenditure of diabetic patients and to investigate whether the outpatient medical resources expenditure of diabetic patients had been significantly lowered by the PBQ system. Furthermore, it aimed to analyze the influences of the characteristics of the patients and hospitals on the outpatient medical resources expenditure of the diabetic patients. The research was a cohort study which analyzed the data of the outpatient medical resources expenditure of diabetic patients in 2004. The control group was randomly selected from the individuals who did not join the PBQ system. The number of samples was determined by the demographic characteristics of the experimental group, consisted by the individuals who had joined the PBQ system. Based on such criteria, 273,738 samples were selected from the population. After the sampling, the data were analyzed by the independent sample t-test and multiple regression analysis. The main findings of this research were as followed: 1. The PBQ system did restrain the outpatient medical resources expenditure of diabetic patients. 2. Different characteristics of both the patients and hospitals indeed had significant effects on the outpatient medical resources expenditure of diabetic patients. However, there were certain limitations of the research. The samples were all selected from the data in 2004 which incurred a less reliable time frame. In addition, merely the independent variables regarding outpatient medical resources expenditure were used in the study. Therefore, it is suggested that future studies should extend the research period and consider other influential elements, such as the inpatient data and the clinic medical care quality indicators, which could provide more insightful and complete findings of the PBQ implementation in terms of diabetic care to both the researchers and practitioners.