It is difficult to monitor pitch bearings with methods based on vibration signal or strain. In light of this limit， a method based on the acoustic emission （AE） technology is proposed to obtain signals， and the short time Fourier transform （STFT） method is introduced to analyze and diagnose the signals. First， the AE technology is studied， and the principle and process of STFT are deduced. Then， the experiments are carried out on a wind turbine pitch bearing. The data of pitch bearing with and without cracks are compared in time domain， frequency domain and time-frequency domain. It shows that the STFT analysis method works best in detecting the cracks in time-frequency domain. Finally， the characteristics of cracks are determined by new data. The results show that AE signals perform better in grasping the state information of pitch bearing， and STFT analysis method is prior in identifying crack faults. The STFT method maintains its performance when the working conditions or other factors change， which is more practical.