An active suspension design to improve lateral ride quality and stability in a track/vehicle system subject to lateral track irregularity is presented. The measurement of the state variables is performed in a noisy environment, and unknown state variables of the system are estimated from the measurement data by using a Kalman filter. It is assumed that the lateral track irregularity and measurement noise are Gaussian random processes, respectively. The optimal control for the active suspension is determined by minimizing the quadratic performance index composed of the state variables and control efforts, and then the active suspension structure has a cascade feedback loop composed of the Kalman filter and the optimal controller. The numerical results indicate that the proposed active suspension provides much improved lateral ride quality and stability.