Radar rainfall estimates have been exploited and widely used in the world for hydrological simulations thanks to the advantages of spatial distribution compared to observed rainfall at ground stations. In Vietnam, due to the weather radar system has been upgraded and installed in recent years, the exploitation and use for hydrology is still very limited. Because radar rainfall estimates often consist of many errors arising from various sources, before using for flow simulation and prediction, bias correction is very necessary to minize errors. Therefore, this study was conducted to correct the radar rainfall estimates in order to improve the input quality for runoff simulation in Vietnam. Two methods including climatology and Kalman filtering technique are applied consecutively to increase the accuracy of radar rainfall estimates. The results of applying to the Ma River basin show that the quality of radar rainfall estimates has increased significantly with the improvement of mean error, correlation coefficicent and root mean square error.