Abstract: The ground-based multi-channels microwave radiometer can provide high temporal resolution profiles of the atmospheric temperature, humidity and liquid water and has a potential application in investigating the formation and evolution of cloud and fog weather events. However, the verification studies of the data retrieved from the multi-channels microwave radiometer are few. In this paper, the reliability of the microwave radiometer was tested by contrasting the brightness temperature simulated by the MonoRTM with that observed by the instrument. The temperature (T), water vapor density (ρ) and relative humidity (RH) data retrieved from the Microwave Radiometer Profiles (MWRP) with 35 channels produced by the Radiometrics Corporation were examined by the radiosonde data of 170 times in which 30 times are in fog periods covering the 11 fog events during 2009-2013 in northern China. The tethered balloon sounding and CloudSat cloud radar data were also used to examine the temperature and humidity profiles derived from MWRP for some typical fog events. The comparison of MWRP's profiles with radiosonde soundings shows that the correlation between two observational methods in terms of temperature data was higher than 0.98, and that in water vapor density data reached more than 0.95. The correlation between these two relative humidity data was much lower, only about 0.67. In the error analysis of the overall samples, T retrieved from MWRP was lower about 3℃ than that from radiosonde sounding, the mean root mean square error (E) of ρ was less than 1 g/m3, and RH from MWRP was larger than that from radiosonde sounding between the height of 1 and 7 km. The mean E of RH was 18% for overall samples and 23% for foggy samples. The comparison of profiles between the tethered balloon soundings and MWRP shows that the retrieved T from MWRP was also lower than that by tethered balloon, but both were in good agreement in indicating the development and evolution of fog events. The high RH at upper-levels retrieved from MWRP was relevant to the clouds as shown by the cloud radar data from CloudSat.