Abstract: Based on adaptive neuro- fuzzy inference system, the adaptive neuro- fuzzy fault diagnosis model of mine hoist was constructed, taking current signal, p ressure signal from hydraulic station, hoist load, hoist speed and accelerated velocity signal collected from hoist system as input variable. This adaptive neuro- fuzzy fa ult diagnosis was based on subtraction clustering algorithm and parameters of machine, electric and hydraulic pressure in hoist system were inducted into adaptive neu ro- fuzzy fault diagnosis system after properly processed, as the input vectors of adaptive neuro- fuzzy fault diagnosis model. Via hoist operating data collected from ma in shaft hoist system, adaptive neuro- fuzzy fault diagnosis model was trained. After the successful training, this prototype system was adopted successfully to diagnose the situation of overloading, transferring heavy objects to a lower level and hydraulic station under pressure in hoist system, verifying the effectiveness of this diagnosis strategy.