To increase the accuracy of radar rainfall estimation, an automatic convective/stratiform precipitation classification algorithm based on radar reflectivity data at the Wuhan radar station is proposed. Based on the two-dimensional reflectivity field, it can identify bright bands, extract the information about the three-dimensional structure of precipitation, and use them to correct the preliminarily analyzed results. Evaluation shows that the method can separate most of the convective and stratiform cloud precipitation correctly, and the performance is obviously better than that by using two-dimensional information. Especially it can identify stratiform precipitation with an obvious bright band and weak echoes in the periphery of the convective core.