Wavelet bicoherence is considered a state-of-the art technique in detecting the phase coupling and nonlinear interactions that occur in a signal or signals. It is used in this study to analyze the nonlinear wind-wave interactions for the simultaneously measured wind speed and wave height during 6 hours of wave generation in a mistral event. The results show that wavelet coherence can detect the linear coherence between wind and wave. On the other hand, the wavelet bicoherence has the ability to detect the phase coupling and nonlinear development between wind and wave during wave growth. This study shows that the phase coupling and nonlinear wind-wave interaction can occur over a certain range of frequencies and that this range is different from one record to another because of the nonstationary feature of the data set.