This paper discusses an adaptive dynamic surface control method for uncertain strict feedback nonlinear systems with input saturation and external disturbance. The radial basis neural network (RBFNN) is used to approximate the unknown function and the hyperbolic tangent function is used to solve the input saturation problem. Meanwhile, the improved dynamic surface control technology reduces the influence of first-order filter on the state error, and the control scheme has excellent tracking performance and stability. Finally, the final bounded convergence of all closed-loop signals is proved, and the superiority of the proposed control scheme is verified by simulation.