Abstract: In order to reduce the probability of fault occurrence of local ventilation system in coal mine and prevent gas from exceeding the standard limit, an approach incorporating the reliability analysis, rough set theory, genetic algorithm(GA), and intelligent decision support system (IDSS) are used to establish and develop a fault diagnosis system of local ventilation in coal mine．Fault tree model is established and its reliability analysis is performed．The algorithms and software of key fault symptom and fault diagnosis rule acquiring are also analyzed and developed. Finally, a prototype system is developed and demonstrated by a mine instance.The research results indicate that the proposed approach in this paper can accurately and quickly find the fault reason in a local ventilation system of coal mine and can reduce difficulty of the fault diagnosis of the local ventilation system.