Sounding observation of humidity is an important means of obtaining the atmospheric humidity data. The measurement results of humidity and accuracy of cloud recognition based on that are directly affected by the performance of the radiosonde humidity sensor. In order to better understand the accuracy of the current measurement of the radiosonde at home and abroad, a large number of research results are reviewed. The types of radiosonde and the performance of its humidity sensor are briefly introduced. Moreover, the influence of humidity measurement error on cloud recognition is also discussed. The results show that the error sources of radiosonde humidity measurement are various and it’s a comprehensive result of many factors. In general, accuracy of humidity measurement is more reliable in the low troposphere with high temperature and so with the cloud identification by the humidity. However, the response time is longer and sensitivity of humidity sensor is lower at low temperature, which results in the accuracy of cloud bottom recognition being higher than that of cloud top while the medium and high cloud recognized by radiosonde being less than the reality. Moreover, under high-humidity conditions, the humidity sensors are easily wetted, which leads to the abnormally high value of humidity and resulting in thicker cloud. Furthermore, the radiosonde generally has low humidity anomalies, especially when the synoptic station with high humidity, resulting in missing report of cloud.