Vibration isolation is almost not taken into consideration in micro-cultivator design at present. When the machine is running, it will generate excessive vibration and make the operator unbearable. Aiming at solving the problem, the mathematical model of magnetorheological (MR) elastomer with variable stiffness and damping was established based on the model of MR fluid and the equivalent stiffness and damping were calculated by the model. Meanwhile, a new structure MR elastomer vibration isolator was designed and mathematical model of vibration isolation system was built. The influence of stiffness and damping to the system stability was studied then. Subsequently, variable stiffness and damping control strategies were constructed according to the characteristics of MR elastomer. Based on these data, the Simulink simulation model was constructed to carry out simulation analysis to verify the control effect from the simulation results by adopting PID control, ON/OFF control and no control strategies. The results showed that: in step input, the displacement and velocity responses of PID control system would tend to stability much more rapidly than those of ON/OFF control. The maximum displacement and velocity amplitudes of ON/OFF control were about two times and eight times than those of PID control, respectively; it would take a very long time for displacement and velocity responses of no control tending to stability and the amplitudes were the largest. In multi-frequency sine input, the displacement amplitude of no control was five times more than that of ON/ OFF control and about ten times than that of PID control. The velocity amplitude of no control was seven times than that of ON/OFF control. However, the velocity amplitude of ON/OFF control was slightly larger than that of PID control. In random input, the displacement amplitude of ON/OFF control was smaller than that of no control, but it was larger than that of PID control overall. Furthermore, the displacement curve of PID control was the smoothest. The velocity curve fluctuation of PID control was the smallest, the ON/OFF control was the second, and the fluctuation without control was the largest. In hybrid input, the displacement and velocity amplitudes of no control were about four times and seven times than those of ON/OFF control, respectively and the displacement and velocity amplitudes of ON/OFF control were both two times than those of PID control, respectively. The displacement and velocity curves fluctuation of PID control were the smallest. In conclusion, PID control strategy is simple and PID control effect is the best. It will lay a theoretical foundation for future controller development and experiment verification.