Effective and objective assessment of the high-resolution numerical model ability to forecast gridded precipitation is not only the basis for the development of the model, but also directly related to the main objective of China Meteorological Administration to improve gridded weather forecast. Based on the ECMWF (European Centre for Medium-Range Weather Forecasts) forecast of high resolution gridded precipitation data, CMORPH (NOAA Climate Prediction Center Morphing Method) satellite data, and the fusion of hourly precipitation data collected at 30000 automatic weather stations in China, 55 precipitation cases of high resolution gridded precipitation forecast from June 2015 to August 2015 were selected for verification. Advantages and disadvantages of the traditional method were compared to that of the object-oriented MODE (Method for Object-based Diagnostic Evaluation) method and the neighborhood method to provide references for the evaluation of high-resolution gridded precipitation forecasting. The main results are:(1) Although the traditional point to point method has some limitations in verifying high-resolution gridded precipitation forecast, the traditional method can reveal regional differences in high-resolution gridded precipitation forecast skill, and depict the overall temporal performance of the model forecast. Thereby it still has important practical value for performance evaluation of high resolution gridded forecasts. (2) one significant advantage of the neighborhood method is that it can obtain traditional forecasting techniques on different spatial scales by varying the neighborhood window. On the other hand, the unique skill of this method to obtain the FSS (Fractions Skill Score) score can provide the ratio of the number of points with precipitation between observed and predicted precipitation. Combining the traditional techniques with neighborhood FSS scores on different neighborhood windows, the spatial scale on which the model has a better performance of forecast can be determined. (3) In the MODE precipitation verification method, an object is extracted on the basis of conversion convolution radius. Such objects can not only provide information similar to that obtained by the traditional techniques such as forecast skill scores and scale-depending forecast performance, but also show the centroid distance between precipitation objects, the axis angle, size, strength, comprehensive income, displacement distance, etc. These attributes can provide information of the model forecast performance from multiple perspectives and quantify errors of the weather system development speed, strength and other forecast variables. Thereby this method has unique advantages. However, there still exist some difficulties regarding how to improve the practical forecasting capability using these object attributes.