Abstract: Shortcomings exist in the traditional cloud model which is unable to discuss the certainty and uncertainty relationships among indicators of interval distribution as a whole. Based on the characteristics of rock mass classification index distributions and fuzzy classification boundaries, a method of finite interval cloud model and distance discrimination is proposed. The calculation model and concept of finite interval cloud model and related basic theory are described, and the classification index weights are obtained by the distance discrimination method based on the evaluation index system and their classification criteria. The cloud of each indictor at different levels is generated according to the classification indexes. Considering the measured index values and the corresponding weights, the comprehensive degrees of certainty are obtained and the rock mass level is identified by the maximum subordination principle. An example is introduced to further explain the calculation flow of the classification model. Comparison with other theoretic methods shows that the proposed model is feasible and effective for dam rock mass classification, and it provides a novel idea for similar problems.