Induction motors are widely preferred in industrial applications to provide mechanical energy. Stator winding and bearing faults are the most common fault types that must be handled on condition monitoring of induction motors. Condition monitoring of single fault of induction motor can be effectively monitored by applying conventional Motor Vibration Signature Analysis (MVSA). However, the effectiveness of these conventional methods is limited in the presence of multiple faults due to the increasing number of harmonics under varying loading conditions. This paper focuses on the detection of characteristic harmonic components of simultaneous multiple faults including stator inter-turn short circuit and outer-race or inner-race bearing faults under four loading levels based on vibration signal analysis by applying Hilbert envelope analysis. The experimental results show that the proposed method can effectively detect fault characteristic harmonic under 25%, 50%, 75% and 100% loading levels of induction motor.