The turbulent momentum fluxes through the air-sea interface are usually measured by the eddy covariance method, in which the measurement plane must be parallel to the horizontal plane. In practice, it is impossible to exactly satisfy this requirement, thus tilt corrections are needed. Three tilt correction methods, including double rotation, triple rotation, and planar fit, have been proposed. This study comprehensively investigates the influence of tilt correction methods on the air-sea momentum flux by utilizing observational data from an offshore platform in the South China Sea. It is shown that the momentum flux corrected by the double rotation method is consistent with that from the planar fit method in the condition of neutral stratification of the atmosphere. In the condition of stable stratification, the planar fit method demonstrates the best adaptability. On the other hand, the triple rotation method is failed to further improve flux on the basis of the double rotation method. It is also indicated that the effects of various tilt correction methods can be neglected on the turbulent kinetic energy spectrum, wind stress direction, and the angle between wind stress direction and wind direction. The wind stress directions are on the left, consistent and scattered with the wind directions under stable, neutral, and unstable stratification of the atmosphere, respectively. However, the drag coefficient decreases under stable conditions. The relationship between the drag coefficient and wind speed from this study generally agrees with those of the previous studies.