This paper focuses the trajectory tracking control of switched robot manipulators systems by adaptive neural network method. According to the different working condition of robot manipulators, the dynamics of robot manipulators is expressed as a switched system with two subsystems. One subsystem is unloaded subsystem and the other is loaded subsystem. Two neural network controllers are designed respectively, and the switching law is the dwell time switching signal. Stability of systems are proved by the multiple Lyapunov functions method. Simulation results show the proposed method can realize the trajectory tracking of the robot manipulators accurately.