In order to improve the performance and the stability of the sensorless vector control system for permanent magnet synchronous motor, H∞ speed observer of permanent magnet synchronous motor based on Immune Algorithm (IA) is proposed in this paper. The H∞ speed observer is used instead of speed sensor to implement the speed identification for permanent magnet synchronous motor. The problem that the weighting factor of the observer is difficult to be selected restricts its application of high-performance in the variable frequency system. So the IA algorithm is introduced into the H∞ speed observer, and the observer coefficient is optimized by taking full advantage of the adaptability of IA algorithm. Therefore, the performance of the sensorless vector control system for permanent magnet synchronous motor is improved better. The simulation and experimental results verify the correctness and effectiveness of the algorithm.