This study developed a bus scheduling expert system to reschedule diffe rent bus routes under those abnormal circumstances such as the absence of bus drivers, breakdown of vehicles, increase of boarding demand, as well as traffic congestion. The most important contents in this system include a bus scheduling simulation model and a knowledge base. The simulation model was used to generate normal schedule with the objective of minimizing total social cost. The knowledge base was constructed by those knowledge obtained from visting experts and field studies. To deal with abnormal circumstances of bus operations, the best alternative of mutual support can be chosen from the base for rescheduling.