The principle and modeling method of movable statistical diagnostic model of soil moisture based on soil initial water content and time period daily average precipitation were introduced. Model was established by the data of 87 monitoring sites in 23 counties in 7 provinces during 2012-2014, and validated by the data of 2015. The movable statistical diagnostic model includes time interval model and daily time series model. The results showed that the qualification rate of the time interval model prediction was about 90%, and the qualification rate of daily time series model prediction was about 70%, indicating that time interval model was better than daily time series model. movable statistical diagnostic model needs to be further improved as a soil moisture diagnostic model. The soil initial water content and precipitation should be segmented separately and then modeled respectively with other models.