In order to ensure the high dynamic and accurate tracking of the desired trajectory, an optimal trajectory tracking control strategy with specified performance based on reinforcement learning was proposed for the an unmanned surface vehicle (USV) trajectory tracking system with input saturation limit. The hyperbolic tangent function was used to approximate the saturation function, and by the specified performance control technique, the tracking error of the unmanned ship was transformed, so that the tracking dynamic error was constrained within the expected specified range to ensure the high dynamic performance of the trajectory tracking system. On this basis, the optimal control theory and the actor critical control framework based on neural network in reinforcement learning were used to design the optimal control method of unmanned ship trajectory tracking, and the problem of control input saturation limit was solved. Based on the Lyapunov stability theory, the stability of the proposed trajectory tracking system under the control strategy was proved, and the tracking error was guaranteed within the specified range. Simulation results also verify the effectiveness and superiority of the proposed strategy.