The fault of broken rotor bars in a three-phase induction motor is diagnosed by using the wavelet multi-resolution analysis in this paper. Daubechies wavelet is selected as the wavelet base, and the wavelet coefficient is obtained from the wavelet transform of the current signal of the faulty induction motor. The wavelet reconstruction of a single branch is then processed. This is the signal in each frequency branch as a function of time. A rotor fault in the induction motor is clearly identified from the distinguishing current characteristics on the high frequency branch, compared with the no fault condition.