The key work of this paper proposes a fuzzy PID control scheme to automatic vehicle obstacle avoidance. By the genetic algorithm, the local optimum parameters are solved. The fuzzy rule is used to tune intelligently the optimal parameters among some range. On the basis of relationship of inputs and output, a fuzzy inference rule is made up. To demonstrate the effectiveness of the proposed rule, a prototype board is developed in this paper. A PC based operating platform is used to have human interface characteristic and fast process capability. Finally, the experiment will show robustness and smoothness of the fuzzy PID rule about that automatic vehicle obstacle avoidance.