A list control method is proposed, which considers the acquisition rate, acquisition punctuality rate, quality control accuracy rate and model consistency rate of observation data. The wind observation data of 120 radiosonde stations in 2019 are used to validate the method, and analyze the quality of observation data. The results are as follows. The stations with observation data problems can be effectively checked out by the list control method. Compared with the model data, the observation data of the list stations have obvious systematic deviation, and the bias and root mean square error (RMSE) are significantly larger than the national averages. The quality of radiosonde wind observation data is good, and the wind direction and wind speed observation data are relatively consistent with the model data in all the four seasons with the biases are within ±1° and ±1.5 m·s-1, respectively. The wind direction consistency in autumn is the highest, and the wind speed consistency in summer and winter is lower than that in spring and autumn. The wind direction consistency first decreases and then increases with the decrease of pressure in spring and summer, while it is opposite in autumn and winter. The wind speed consistency basically changes in a threepeak pattern with the decrease of pressure.