In view of problem of invalidation in dealing with conflicting evidence of D-S evidence theory used in health diagnosis method of coal mine belt conveyor based on multi-sensor information fusion, a health diagnosis method of coal mine belt conveyor based on fuzzy evidence theory was proposed. Firstly, information of belt conveyor are collected by use of multi-sensor, and basic probability assignments are obtained according to membership function, so as to extract information feature. Then information fusion based on fuzzy evidence theory is realized by modifying conflicting evidence and applying synthesis rule of D-S evidence theory. Finally, running state of belt conveyor is judged according to decision rule. Effectiveness of the method is verified by an example.