Abstract: To understand the accuracy and usability of wind profile radar data in complex mountainous terrain, wind profile radar data and L-band sounding radar wind data at Guiyang station from January 2023 to June 2024 were selected. A comparative analysis of the two datasets was conducted under the different altitudes, different observation times, and under clear versus rainy weather by calculating the mean error, standard deviation, and effective sample rate for the u and v wind components. The results showed that: (1) The wind profile radar and sounding wind measurements results had good consistency, with standard deviations of 2.43 m·s⁻¹ and 2.68 m·s⁻¹ for the u-component and v-component respectively. The u-component measured by the wind profile radar was smaller than that from the L-band sounding radar, while the v-component measured by the wind profile radar was slightly larger. (2) Compared with the L-band sounding, the u-component near the ground measured by the wind profile radar was larger and the v-component was smaller, while the u-component in the middle-to-upper level measured by the wind profile radar was smaller and the v-component was larger. The standard deviation of u -component at various heights ranged from 1.95 to 2.74 m·s⁻¹, and the standard deviation of v-component below 7 km ranged from 1.67 to 2.96 m·s⁻¹. (3) The standard deviations of u-component and v-component at 08:00 were 2.42 m·s⁻¹ and 2.64 m·s⁻¹, respectively, while those were 2.43 m·s⁻¹ and 2.72 m·s⁻¹ at 20:00. The deviation between the two types of data at 20:00 was larger than that at 08:00. (4) Under precipitation conditions, the standard deviations of u-component and v-component were 2.54 m·s⁻¹ and 2.94 m·s⁻¹, respectively, whereas under non-precipitation conditions those were 2.41 m·s⁻¹ and 2.64 m·s⁻¹. Standard deviations of u-component and v-component during precipitation were both larger than those without precipitation, it can be seen that precipitation will have a certain impact on the data quality of the wind profile radar.