Based on the data of sea surface wind field at 10 m from ECMWF (European Center for MediumRange Weather Forecasts) from 1979 to 2017, the Empirical Orthogonal Function (EOF) method and wavelet timefrequency analysis are used to analyze the variation characteristics of nearsurface wind field over the South China Sea and its response to ENSO. The results are as follows. For the first EOF mode of the nearsurface wind field over the South China Sea, the average sea surface wind speed presents downtrend and interdecadal variations and is correlated with ENSO, but the correlation gradually decreases since 1990. For the second EOF mode, the average wind speed shows downtrend over the northern and southern South China Sea and uptrend over the central South China Sea. For the third EOF mode, the average sea surface wind speed gradually decreases over the central South China Sea and increases over the northern and southern South China Sea. Interdecadal variations are manifested in the second and third mode and are significantly correlated with ENSO. Recent years have witnessed the correlation increase between ENSO and the third mode. In spring, the average sea surface wind speed over the South China Sea gradually increases from south to north. In summer, there is a center of maximum wind speed over some sea areas of the coast of Vietnam, while the speed gradually decreases from these sea areas to the surrounding areas. Southwesterly winds prevail over these sea areas. In autumn, the speed successively increases from south to north. In winter, the wind speed is relatively large over the whole South China Sea, and there are two centers of maximum speed over the coast of Vietnam and the sea areas of Dongsha Islands.