This paper introduces the application of neural network in the motion control of unmanned vehicles in recent years. Due to the non-linearity, time delay and parameter uncertainty of unmanned vehicles, as well as the influence of external factors such as road adhesion coefficient and lateral wind, it is often difficult to guarantee the real-time and high precision of the control by traditional control methods. Neural network has the ability of approximating complex nonlinear mapping, which can adapt to the environment of unmanned vehicle operation and complete motion control. This paper introduces the application of neural network in motion controller design of unmanned vehicle, accurate approximation of unknown dynamic parts of the system, estimation of vehicle state parameters and parameter optimization of other control algorithms, and gives the main content and latest progress of related research. Finally, the main problems of neural network in the motion control of unmanned vehicles are summarized, and the possible development direction is prospected.