For solving the problems of stochastic disturbance and input saturation existing in permanent magnet synchronous motors (PMSMs) drive systems, a command filter-based adaptive neural control method is proposed in this paper. Firstly, the neural networks technique is utilised to approximate unknown nonlinear functions. Then, the command filtered controller is constructed to avoid the 'explosion of complexity' inherent in the classic backstepping control and the error compensation mechanism is introduced to reduce the error caused by command filter. Moreover, the adaptive backstepping method is utilised to design controllers to assure that all signals are bounded in the closed-loop systems. Finally, the effectiveness of the approach is certified by the given simulation results.