Abstract: The polarimetric radar is an important detection device whose measurements can be used for severe convective weather analysis and cloud microphysics progress research. Upgrading the traditional Doppler weather radar to polarimetric radar is a key part of severe convective weather monitoring program of China in the next few years, and the quality control of polarimetric radar measurements is key technical issue of the monitoring program. In Guangdong Province, based on the domestic and international mainstream quality control algorithms and relevant experience, a quality control system is developed for S-band polarimetric radars, to deal with the non-meteorological echo, non-standard blockage and high frequency noise in the radar radial, which have negative impacts on application of polarimetric radar measurements in data assimilation. The system is applied to the typical severe convective weather case in South China monsoon region, including a rainfall case, a severe convection case and a typhoon case in 2017. Evaluation results show that a combination of the hydrometeor classification screening based on fuzzy logic, co-polar cross-correlation coefficient (ρ_HV), signal-to-noise ratio (SNR) and specific differential phase (K_DP) thresholding and despeckling can remove most non-meteorological echoes, and suppress virtual echo caused by anomalous propagation efficiently. Non-meteorological echoes include ground clutter, biological scatters, partial clear-air echo and radiographic noise due to anomalous propagation. A linear interpolation is employed to fill the small gap (the width of which is less than 5°) caused by non-standard blockage. A median filter and radial smooth are found effective in filtering out high frequency noise in the radar radial while maintaining polarimetric radar characteristics. After quality control, the meteorological echo is clearer and more prominent, and accounts for about 40% of valid observation which is defined by reflectivity (Z_H) being larger than -30 dBZ. Z_H of the meteorological echo is larger than 5 dBZ, ρ_HV is larger than 0.8 and less than 1.0, and the differential reflectivity (Z_DR) is between -0.2 and 4 dB. Batch tests are needed to keep the quality control system stable and effective in the further work. And how to combine multiple polarimetric radar measurements to form a three-dimensional gridded product is also another important prerequisite for application of polarimetric radars measurements in the numerical model.