A comparative study of sliding mode and model reference adaptive speed observers for a speed sensorless induction motor (IM) drive is introduced in this paper. First, a speed observer on the basis of the sliding mode control theory is proposed. Then, the sliding mode speed observer is compared with the speed estimation algorithm based on model reference adaptive system (MRAS) theory to verify the robustness against parameter variations. An integral proportional (IP) speed controller using the estimated speed signal is designed to stabilize the speed loop. Moreover, to increase the accuracy of the estimated speed, the speed estimation algorithms are implemented using a digital signal processor (DSP). The effectiveness of the proposed sliding mode speed observer is demonstrated by some simulation and experimental results.