The continuous wavelet transform enables one to look at the evolution in the time frequency joint representation plane. This advantage makes it very suitable for the detection of singularity generated by localized defects in mechanical system. This gives a desirable ability to detect the singularity characteristic of a signal precisely. In this study, the impulse response wavelet time frequency joint representation is used in conjunction to detect and diagnose the bearing fault. The impulse response wavelet time frequency joint representation is found to show distinctive signatures in the presence of bearing inner race or outer race damage. The experimental results show that the impulse response wavelet time frequency representation can extract the transients from strong noise signals and can effectively diagnose bearing fault.