Abstract: A neural network sliding mode control method based on switching gain regulation is proposed for the position tracking of robot. Firstly, a sliding mode controller was designed for the robot. Then, the switching gain of the sliding mode controller was regulated by neural network. The switching gain can be changed with the changing of uncertainties such as external interference. Thus, the switching gain can be estimated in real time and the chattering problem in the traditional sliding mode control can be solved. Finally, the control method was verified on a double-arm robot. The simulation results show that the robot can track the desired trajectory better and alleviate the chattering problem more effectively with the proposed neural network sliding mode control method than the traditional sliding mode control.