Aiming at the different index selection and requiring a large amount of prior knowledge in the comprehensive reliability evaluation of CNC machine tools,an evaluation method combining rough set with K-means clustering algorithm was proposed.The original evaluation index system was established,each index was discretized through the K-means clustering algorithm and Silhouetta index,and the rough set theory was used to reduce the redundant index,to construct a dynamic evaluation index system.According to the definition of attribute importance,objective weight was given to the reduced index,a comprehensive reliability evaluation model of rough set-clustering was constructed.The results show that the proposed method is reasonable and effective.