The principle and method of integrated diagnostic model of soil moisture were introduced, which synthesizes balance diagnostic model, statistical diagnostic model, subtractive statistical diagnostic model, ratio statistical diagnostic model, interval days statistics diagnostic model and moving statistics diagnostic model. The model was established by using the data of 87 monitoring sites in 23 counties in 7 provinces during the period of 2012-2014, and was validated by using the data of 2015. The results showed that the integrated model had high qualification rate (>97%) in diagnosis and prediction. The advantages of integrated diagnostic model are embodied as: (1) the parameters of the six independent models are different under different climates, underlying surface and management measures. Therefore, the six independent models are not universal even by sitespecific statistics. (2) In the same prediction, one of the prediction results of the six models always has relatively high accuracy. This result is highly correlated with the sum of the precipitation in different periods and the initial soil water content. After adaptive optimization of the six models, the predicted result is selected according to the sum of water income by the probability.