speed tracking capability of model reference adaptive system (MRAS) with model-based flux/speed observers and artificial neural network (ANN)-based adaptive speed estimators for sensorless induction motor (IM) drives has been analyzed. In model-based technique, mathematical model of IM is used to estimate the rotor speed. The current and flux observers are used as the reference model to estimate the rotor flux. The estimated rotor flux signals are used as the input signal for the adaptive observer to estimate the speed. In ANN-based method, adaptive model is constructed with a feedforward neural network to estimate the rotor speed. Feedforward ANN algorithm is used to train the network. The training algorithm decides the learning speed, stability, and dynamic performance of the system. Both methods have good speed tracking capability. Simulation results are presented to know the accuracy of the proposed methods. The proposed speed estimation techniques have great potential in industrial applications.