Abstract: Induction motors, especially squirrel cage motors, have an important role in industry. Their rotor or stator may be failed under stresses depending on their application and sometimes an unexpected motor failure in a manufacture production process may result in unexpected and unforeseen tripping. So, if the failure can be detected during its operation, there are some methods to prevent failure spread and also manufacture trip. This study aimed to model and simulate the squirrel cage induction motor at any operation condition such as healthy condition and broken bars failures in side of rotor, in order to achieve an algorithm for failure detection. So, in this study a new method base on Wavelet theory is used for failure detection. Finally, simulation and experimental results are compared. The results show that using wavelet transform can effectively be used to diagnose the broken bars in the motor.