The invention provides a river water quality monitoring data sequence encryption method. Firstly a regression equation between a pollutant flux and flow is optimized through an LOADEST model, a water quality sequence encrypted for the first time is pre-estimated through high-frequency flow monitoring data, and a fluctuation characteristic of a flow process is transmitted to a water quality process through the regression equation, so that the deficiency that a linear interpolation method cannot describe a fluctuation characteristic of a water quality change process is made up for; secondly the water quality sequence encrypted for the first time is corrected through a Kalman filtering correction model, and a water quality actual measurement value and a water quality estimation value based on the LOADEST model are subjected to data assimilation, so that an error between the final estimation value and the actual measurement value can be effectively reduced and the deficiency of relatively large error of related model interpolation methods is made up for; and finally the obtained encrypted water quality data sequence can not only describe the fluctuation change characteristic of the water quality process between two actual measurement points but also effectively reduce the error between the estimation value and the actual measurement value, so that the defects of an existing method are overcome.