One of the useful methods for rock slope stability analysis is the Romana SMR classification. This method is a developed version of the Bieniawski's 'rock mass rating' (RMR) system. This classification is based on classic set theory. Characterization of rock's mass is very complex and may result in some ambiguous. The classic Sets Theory Classification is not able to yield to unambiguous results. Using fuzzy set theory is an effective approach to quantify these ambiguities. This paper describes the application of fuzzy set theory to SMR classification by incorporating fuzzy sets. In the proposed approach the Mamdani fuzzy algorithm was constructed using 825 ''if-then'' rules for evaluating rock slope stability. In addition, slope instabilities in an open pit mine were tested and results were evaluated to confirm the accuracy of implementation this proposed approach.