Abstract: This paper describes the design of a prescribed performance for the piezoactuator-driven stage with control input saturation and hysteretic nonlinearity by using the backstepping scheme. Neural networks are used to approximate the hysteretic nonlinearity of the system and compensate for input constraints. A system-equivalent error model is established using the prescribed performance function and error transformation function. Taylor polynomials are used to express a new equivalent error, and the singular value problem in the conventional prescribed performance control is avoided. The controller of the entire system is designed step by step by using the backstepping control method. Lyapunov functions are used to prove that the signals of the entire closed-loop system are uniformly ultimately bounded, and the output error is limited to the prescribed range. Simulation results show the effectiveness of the method.