Abstract: In order to improve the comfort of the vehicle, the research on magnetorheological semi-active suspension control considering time delay is carried out. Firstly, according to the robust control theory, the H_∞ control method which can obtain the desired control force of the suspension was worked out. Then, the inverse model which can reflect the relationship between the desired control force and the control voltage was established, and the active control to semi-active control transition was realized. In the control strategy, for the time-delay problem of the control system, the control method under the two stability conditions of time-delay independence and time-delay dependence was obtained respectively. It is difficult to establish an inverse model of the analytical form for the magnetorheological damper. The BP neural network was utilized to setup the relationship between the desired control force and the control voltage. Finally, the semi-active suspension control performance verification was carried out. The results show that the suspension control performance is conservative under time-delay independent conditions, but the stability is high. The suspension control performance under time-delay dependent conditions is better than that under the time-delay independent conditions, but the time-delay boundary has great influence on the control performance.