To solve the problem of difficulty in obtaining the parameters of the dynamic model of underactuated unmanned surface vehicle, and the tracking control is susceptible to interference from water flow and natural wind, this paper proposes a trajectory tracking control method based on adaptive time delay estimation and super-twisting fast terminal sliding mode. First, the hand position point coordinate transformation method is used to reconstruct the state expansion of the underactuated unmanned surface vehicle mathematical model; then adaptive time delay estimation technology is used to estimate the unknown dynamic model part of the unmanned surface vehicle and external disturbances, and design a super-twisting fast terminal sliding mode control strategy compensates for errors caused by time delay estimation and improves the robustness of the control system; finally, the feasibility of the control method in this study is verified based on simulation analysis and navigation tracking experiments. In the trajectory tracking experiment, the average positional error was reduced by 35.1%, 24.7%, and 20.8% compared to the PID algorithm, sliding mode algorithm, and adaptive delay estimation sliding mode algorithm, respectively, demonstrating high trajectory tracking accuracy. The control method proposed in this paper does not rely on the parameters of the dynamic model and can achieve precise trajectory tracking control for underactuated unmanned surface vehicles. It features fast convergence speed and high robustness, providing a reference method for trajectory tracking control of underactuated unmanned surface vehicles.