Objective:To analyze the input and output status of health resources in primary medical and health institutions and their allocation efficiency in different regions of China, and to provide an empirical basis for optimizing the allocation of primary medical and health resources in China among regions.Methods:The input index data (number of beds and number of health personnel) and output index data (number of primary medical and health institutions visits, number of family health services, number of hospital admissions) of primary medical and health institutions in China in 2020 were extracted from the China Health Statistical Yearbook 2021. Based on the BCC ( Banker, Charnes, Cooper) model of data envelopment analysis ( DEA), the Bootstrap- DEA method was used to correct bias, the allocation efficiency of primary medical and health resources in 31 provinces was calculated and the regional differences were analyzed. Results:After bias correction, the technical efficiency (TE) of resource allocation in primary medical and health institutions decreased by 0.102. The average TE score of all 31 primary medical and health institutions was 0.669, indicating a serious problem of ineffective use of technology. The TE of the eastern, central and western regions was 0.694, 0.663, and 0.649 respectively. There was obvious polarization in the central regions.Further analysis of the efficiency improvement of non DEA efficient provinces showed that 2 DEA weakly efficient provinces and 16 DEA ineffective provinces had several reference provinces for efficiency configuration improvement; The provinces that have been referenced more than 10 times were Zhejiang, Chongqing, Sichuan, and Ningxia, while the provinces that were listed as the first reference by other provinces were Ningxia, Chongqing, Zhejiang, and Tibet.Conclusions:The resource allocation efficiency of primary medical and health institutions in China is relatively low, and regional differences are obvious. The balance between different inputs and outputs should be considered when allocating the resources. Non DEA effective provinces can use DEA analysis to find the most suitable reference object and make reference improvements in the short term.