Abstract: Model is an important tool for studying the change of water environment and water environment management. The spatially referenced regressions on watershed attributes (SPARROW) is a non-linear watershed regression model based on the mass balance, which correlates the monitoring data with watershed characteristics and pollutant source information. It has the advantages of less data requirements, clear structure and excellent versality. In order to deeply understand the application situation and development trend of the SPARROW model in water environment management, the principles of the SPARROW model and its application in nutrient background concentration simulation, water quality evaluation, water quality target management and the impact of climate change on water environment were systematically reviewed. The results show that: (1) The SPARROW model could effectively simulate the background nutrient flux and concentration by choosing appropriate reference sites, and provide basis for the formulation of water quality standards. (2) The model can extrapolate nutrient concentration of unmonitored area via monitor data, which makes it possible to evaluate water quality under the condition of limited water quality monitoring data. (3) The model can simulate the river nutrient load under different land use patterns and resource management, and provide support for water quality management and decision-making. (4) Under climate change scenarios, the study of the impact of climate change on the water environment based on the SPARROW model can assist the formulation of water environment management plans and tackle the increase of nutrient output caused by climate change in the future. In this paper, the problems existing in the application of the SPARROW model are analyzed and discussed, and it is recommended to strengthen the following aspects of research when applying the SPARROW model in the future: (1) The relevant modules such as permanganate index, COD and ammonia nitrogen simulation modules should be further developed; (2) Combine the SPARROW model with machine learning model to improve the ability of quantifying the parameters of the model, so that the model can be better applied to water quality related studies at different scales and different basins.