In this paper, an alternative approach to speed estimation of brushless AC servomotors is presented. Speed control is realized in the following steps. First, the servomotor was mathematically modelled; the driver system was designed and speed control of the servomotor was accomplished with feedback. Next, a network structure representing the electrical and mechanical properties of the servomotor was built via Artificial Neural Network (ANN) and trained with the results of the first step. The weights obtained from the neural nwork training were inserted into the control algorithm in accordance with the ANN. Finally, speed estimation was achieved based on the ANN. The complex servomotor hardware was highly simplified, the cost was significantly reduced, mechanical durability increased, maintenance need dropped considerably, low inertia was gained and the produced noise was noticeably lowered.