Due to the nonlinearity and time delay uncertainty of magneto-rheologcial (MR) suspension, one of the main challenges in the application of MR technology is to derive an appropriate control algorithm. The main purpose of this study is to propose a new adaptive fuzzy sliding mode control (AFSMC) for MR suspension application. Firstly, an accurate control model of a MR damper is formulated by using the measured experimental data, and the control model considering the time delay uncertainty of the MR damper is also developed. Subsequently, a quarter car model equipped with a MR damper is constructed and the AFSMC is formulated. For comparison purposes, a simple sliding mode controller based on the quarter car model is also designed. At last, the numerical simulation and road test are adopted to validate the proposed control algorithm. The results of the simulation and road test show that the MR suspension system with AFSMC can improve greatly ride comfort and avoid the effects of nonlinearity and time delay uncertainty of the MR damper. Furthermore, the control performance of the proposed control algorithm is superior to the simple sliding mode control algorithm.