Abstract: The software and hardware frameworks for fault diagnosis and state prediction of mechanical equipment were established, and the methods and technologies of the diagnosis and the prediction were discussed in detail. Based on the above, a system of intelligent fault diagnosis, prediction and maintenance(IFDPM) of mechanical equipment was developed. This system has been applied to the analysis of grinders in Beijing Jeep Co., Ltd., and the diagnosis result agrees with the fealty, which shows the practicability of the system.