Abstract: In order to establish a tunnel surrounding rock classification method that is based on the entropy method and cloud model to address the fuzziness and randomness in surrounding rock classification, the classification system is treated as a fuzzy system and six parameters are selected as evaluation indexes. A cloud model with various evaluation indexes corresponding to various surrounding rock grades is generated using a normal cloud generator,objective weights are calculated based on the entropy weight method, a comprehensive certainty degree for the evaluated sample is obtained, and thus the grade of the surrounding rock is determined. The entropy-cloud model, which combines the fuzziness and randomness of the evaluation index and rock mass grades, transforms the evaluation index from a qualitative description to a quantitative value, overcoming the shortcomings of the fixed evaluation pattern of traditional tunnel surrounding rock classification. The surrounding rock of five sections of the Xianghe tunnel was classified by an entropy-cloud model, with the results agreeing with those of the FAHP (fuzzy analytic hierarchy process) method and other methods, demonstrating the feasibility and reliability of this method.