Abstract: Quantitative evaluation of the efficiency of municipal wastewater treatment plants (WTPs) is a problem for the sewage treatment industry. Efficiency assessment contributes to making WTP development policy more scientific and reasonable. For this purpose, under the framework of the total factor productivity (TFP), data envelopment analysis (DEA) was employed to establish an evaluation model of the efficiency of WTPs, including input indices of gross fixed assets, annual operating cost, employment and annual power consumption, and output indices of wastewater treatment capacity and pollutant reduction of BOD₅ and NH₃-N. A total of 74 samples were selected as the empirical analysis objects, which have the same discharge standard and treatment process. An efficiency index was obtained for each plant by means of mathematical programming techniques, aiming to determine the relationship between WTPs' efficiency and their scales, and to analyze the sample input redundancy and output slack using a quantitative method. In total, 20 samples reached their relative optimal efficiencies, which means these samples could be benchmarks for other samples. There were 10 samples with pure technical efficiencies, but no scale efficiencies which are needed to improve their scale. 85.2% of the DEA invalid samples showed increasing returns to scale, which indicates the sewage treatment industry in China is in the process of rapid development, with high potential to increase the profits by augmenting investments in the industry. Furthermore, through the Kruskal-Wallis Test, it was determined that the larger plants ran more efficiently than the small plants, which indicates the scale effect in the industry, as was expected. Different levels of input redundancy or output slack existed in the 54 DEA invalid samples; these samples would be the key objects to improve the operating efficiency.