Launched in 2013, and 2017 respectively, Chinese Fengyun-3C and 3D meteorological satellites are equipped with two microwave sounders, Microwave Temperature Sounder (MWTS) and Microwave Humidity Sounder -2 (MWHS-2), whose observations play an important role in numerical weather prediction by data assimilation. Data quality control should be carried out before assimilation to filter out bad data, such as cloud- or rain-polluted data and questionable data. This work can’t be accomplished purely depending on MWTS or MWHS-2 themselves. MWHS-2 is taken as an example to do quality control in the paper, and the method is suitable for MWTS too. Multi-source information from other instruments onboard FY-3 is extracted to assist in the work. Cloud mask product from VIRR (Visible and InfraRed Radiometer), oceanic cloud liquid water content product from MWRI (Microwave Radiation Imager), and global rain rate product from MWRI are mapped to MWHS-2 for quality control in combination with oceanic rain detection product from MWHS-2 itself. 6 kinds of cloud and rain detection schemes are then designed to get the best choice by analyzing the characteristics of background departure. RTTOV v10 is adopted to simulate brightness temperature of MWHS-2 at all channels. The results suggested that scheme RI RC (MWRI cloud and rain information ingested) and scheme RC (all information ingested) are the two best choices for numerical assimilation application, and scheme RI RC can retain more samples. Questionable data can also be found in the way to help monitor the operational status of instruments.