In this paper, the object-oriented methodology is employed to build an environment database for path planning of autonomous guided vehicles (AGVs). The object-oriented environment database (OOED) consists of spatial objects made up of Road, Fork, Region, Obstacle, etc. with non-spatial attributes. An effective path planning algorithm has been presented that comprises two steps: the first is path finding, that can find all possible paths from the start point to the goal by a reasoning method based on the OOED; and the second is path selection that chooses an optimal route by the cost criticism based on human imitation method. The results of simulation have demonstrated that the OOED provides an efficient means for AGV path planning.