Abstract: Aiming at the tracking control problem of hydraulic position servo systems with nonlinear time-varying feature, a design method of filtered-X hydraulic position adaptive inverse control was presented based on adaptive inverse control. A transform domain variable step-size normalized least mean square (NLMS) algorithm was provided to overcome the shortage of traditional adaptive filter algorithms for filtered-X structure. The algorithm was used to in plant modeling, inverse plant modeling and controller design for a hydraulic position adaptive system. Simulation result shows filtered-X hydraulic position adaptive inverse control has the good performances of speediness tracking and strong robust.