Precipitation data is one of the important input data of hydrological model, and it is of great theoretical and practical significance to improve the interpolation accuracy of rainfall data. In order to compare the applicability of different interpolation methods in Ningxia Autonomous Region, based on the average precipitation data of 77 meteorological stations in Ningxia Autonomous Region in 1956-2016, the spatial interpolation and cross-validation of rainfall data by Tyson Polygon, Distance Average Inverse Ratio, Kriging Interpolation method, and the results of method robustness and accuracy are compared and analyzed to arrive at the interpolation method applicable to Ningxia Autonomous Region. The results show that: (1) from the accuracy point of view, the Kriging interpolation method is similar to the direction and position of the observation values and (2) from the point of view of robustness, the average MAE of the Kriging interpolation method is 4.43mm and the RMSE average is 10.36mm, which is significantly smaller than the Tyson polygon method and the distance average inverse ratio. On the whole, the Kriging interpolation method is more ideal and can be used for the spatial interpolation analysis of precipitation in Ningxia Autonomous Region.