Radar-based estimates of rainfall can capture the high spatial variability of the precipitation systems, and it is important for runoff and flood forecast etc. Because of the radar's complexity, radar-based rainfall has errors. This study focuses on evaluating accuracy of the radar-based rainfall data, using METv3.0 software with 1 km×1 km and UW-MEAN space interpolation method. The database includes the CHILL,NEXRAD radar data and ground rain-gauge network data for rain process during Aug. 9, 2008. The result indicates that the METv3.0 software can be used to evaluate the quality of radarbased rainfall data, RZK inversion mode of CHILL radar has the best result and the underestimation and overestimation of the rainfall rate become small by RZK inversion mode. It is important for the further study on the quality of radar-based rainfall data and application of METv3.0 in the analysis of radar-based rainfall data.