Abstract: The regression equations of chemical oxygen demand (COD), total nitrogen (TN) and total phosphorus (TP) fluxes were optimized by LOADEST model based on the discharge and water quality data of Qinjiang river from 2014 to 2019. The R² value for COD, TN, and TP was 0.857, 0.772 and 0.717, respectively. This indicated that the optimized equation could be used to evaluate the variation characteristics of typical pollutants fluxes in Qinjiang river. The results showed: (1) From 2014 to 2019, both concentrations and fluxes of COD, TN, and TP showed a trend of increasing first and then decreasing, and TN still exceeded seriously the water quality standard; (2) The seasonal variation characteristics of COD, TN and TP were significant (p<0.01). The fluxes of COD and TN in wet season were higher than in dry season, indicating that they were strongly driven by rainfall, while their concentrations in dry season were higher than in wet season, suggesting that COD and TN were mainly influenced by non-point sources. Although TP was significant correlated with rainfall, its concentration was also mostly influenced by nonpoint sources. Overall, this study suggested that the construction of sewage pipe network and sewage treatment capacity should be strengthened along the river channel. This will help to curb pollutants input into the river system.