Ocean surface wind vector retrieval research on airborne fully polarimetric SAR (synthetic aperture radar) is of great significance for wind vector sounding under complex weather conditions near the coast. Starting from polarization scattering theory, we designed an ocean surface wind vector retrieval scheme for fully polarimetric SAR, by analyzing the relationship between fully polarimetric SAR sounding data and ocean surface wind vector. According to the high mobility and fully polarimetric two sounding characteristics of airborne SAR, firstly, we presented an ocean surface wind vector retrieval method based on maximum likelihood estimates for VV-polarized sounding data, and designed the flight experiment scheme. Secondly, we proposed an ocean surface wind vector retrieval method for VH-polarized sounding data, retrieved wind speed by VH-polarized ocean surface scattering model from optimum fitting with constraints, and calculated wind direction by CMOD5 geophysical model function. Using typhoon ''Haikui'' edge sounding data from airborne fully polarimetric SAR, we carried out the ocean surface wind vector retrieval experimental research. Results show that two wind vector retrieval methods can retrieve ocean surface wind vector under complex weather conditions without auxiliary information. Wind direction and wind speed retrieval root mean square errors of the former are 18.0°, 1.8 m/s, wind direction and wind speed retrieval root mean square errors of the latter are 9.3°, 1.2 m/s, and the accuracy of the latter is better than that of the former. The VH-polarized normalized radar cross section is more suitable for ocean surface wind vector retrieval under complex weather conditions, because it is independent of wind direction and radar incidence angle but has a linear relationship with respect to wind speed.