Based on the structural health monitoring system (SHMS) installed on Xijiang Bridge of Nanning-Guangzhou Railway, the whole-process data of the wind speeds and directions at the bridge site subjected to Typhoon Kujira in June 2015 were obtained. Utilizing the measured data, the mean wind speeds, mean wind directions, and fluctuating wind parameters including the turbulence intensity, gust factors, turbulence integral scales, power spectrum density (PSD) of fluctuating wind and spatial correlation in different observing points were summarized and statistically analyzed. The analysis results showed that 1-minute mean wind speeds fluctuated in a large margin in the typhoon periphery, and the distribution of measured turbulence intensity exhibited a low discreteness with a longitudinal value of 0.253. Both the turbulence intensity and gust factors decreased with the increase of the mean wind speed. Moreover, large differences between the measured turbulence integral scales and recommended ones in three principal directions were observed. Meso-micro scale vortexes dominated the flow field, whose dispersion slightly increased with the increase of the mean wind speed. Additionally, the Kaimal Spectrum obtained from the relative codes cannot match well the measured horizontal PSD functions, and the measured spectrum value of Typhoon Kujira was greater than those of the codes in both high and low frequency domains. The spatial correlation was apparent in the periphery, and the error was caused when the Davenport assumption was adopted.