X-band-radar-based quantitative precipitation estimation (QPE) system is increasingly gaining interest thanks to its strength in providing high spatial resolution rainfall information for urban hydrological applications. However, prior to such applications, a variety of errors associated with X-band radars are mandatory to be corrected. In general, X-band radar QPE systems are affected by two types of errors: 1) common errors (e.g. mis-calibration, beam blockage, attenuation, non-precipitation clutter, variations in the raindrop size distribution) and 2) “wind drift” errors resulting from non-vertical falling of raindrops. In this study, we first assess the impacts of different corrections of common error using a dataset consisting of one-year reflectivity observations collected at an X-band radar site and a distrometer along with rainfall observations in Beijing urban area. The common error corrections demonstrate promising improvements in the rainfall estimates, even though an underestimate of 24.6% by the radar QPE system in the total accumulated rainfall still exists as compared with gauge observations. The most significant improvement is realized by beam integration correction. The DSD-related corrections (i.e., convective–stratiform classification and local Z - R relationship) also lead to remarkable improvement and highlight the necessity of deriving the localized Z - R relationships for specific rainfall systems. The effectiveness of wind drift correction is then evaluated for a fast-moving case, whose results indicate both the total accumulation and the temporal characteristics of the rainfall estimates can be improved. In conclusion, considerable potential of X-band radar in high-resolution rainfall estimation can be realized by necessary error corrections.