Abstract: In view of the nonlinear characteristics of the semi-active suspension model of vehicles, a study on the application of the adaptive control based on neural networks to semi-active suspension is made. An identifier and a controller with feedforward neural networks are designed, and a specialized learning rule used for the controller is analyzed. The results of simulation show that the ride comfort of a vehicle can be improved by using adaptive control based on neural networks, and the improved BP algorithm used for the neural networks identifier and controller is effective.