Bearing is regarded as a widely used support components in rotating machinery, its failure will seriously affect the safe operation of the equipment. Pointing on the rolling bearing surface damage fault, in this paper the wavelet transform method is used for its fault diagnosis, in this algorithm we can obtain the spectrum of the vibration velocity signal by using the wavelet transform method to decompose and reconstruct the vibration velocity signal and we can get the fault location by spectrum analysis of fault signals. The experimental results show that the wavelet analysis technique can the extract effectively the signals’ characteristic frequency of rolling roller bearing inner ring and the fault, which provides an efficient method for fault diagnosis.