In order to make the vehicle meet the requirements of complex conditions, the dynamic model of 1/2 vehicle magneto-rheological (MR) semi-active suspension (SAS) based on wheelbase preview in steering situation was established, and the variable universe adaptive fuzzy controller based on fuzzy neural network (FNN) was designed. The universe size of the system fuzzy controller can be adjusted automatically by FNN controller according to system deviation, avoided subjectivity of universe division, improved the fuzzy control accuracy. The damping of MR damper can be adjusted by controlling the current size, so the vehicle will remain the ride comfort and safety in steering situation, and the vehicle dynamic performance can be improved. Through the comparison of simulation, the vehicle performance of variable universe FNN control system based on wheelbase preview were superior to passive system and fuzzy control system and without preview system in the same situation.