This paper is concerned with the path following control for an underactuated unmanned surface vehicle (USV) subject to modeling error and input saturation under time‐varying external disturbances. In this note, a novel trajectory linearization control (TLC) method is first introduced in the design of path following controller for an unmanned surface vessel. Adaptive line‐of‐sight (LOS) algorithm is used in the navigation strategy, which not only solves the sideslip angle problem caused by external disturbances but also optimizes the convergence speed. The neural network minimum learning parameter (MLP) method is used to compensate for modeling errors and external disturbances. Compared with multi‐layer neural networks, MLP solves the problem of ‘dimension disaster’. Meanwhile, in order to increase the robustness of the system, adaptive technology is employed to compensate for the compensation error of MLP. In addition, an auxiliary dynamic system is introduced into the controller design to address potential input saturation issue. Considering the above practical conditions, it is more convenient for the engineering implementation of the proposed control path following the control system. Using Lyapunov stability analysis theory, it is proven that all error signals in the system are uniformly ultimately bounded. Finally, two simulation experiments are presented to demonstrate the feasibility of the proposed control strategy. © 2021 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.