Abstract: In view of the uncertainty such as hierarchy and fuzziness of assessing rock slope’s safety and stability, we considered the stability assessment as a fuzzy system and took geological conditions, engineering status, environmental conditions and monitoring performance as well as their influence factors into account, hence building an index system which contains multiple layers including target layer, criterion layer and index layer. Principal component analysis and the improved analytic hierarchy process were combined to present a method which takes into consideration the subjective and objective weights at the same time and improves the accuracy and rationality of index weights. Membership functions were constructed by using fuzzy theory. Finally the comprehensive multi-level fuzzy model for assessing rock slope safety and stability was established. Practical application shows that the model is feasible and easy to operate and the result is in accordance with the reality. The model provides a new approach for the assessment of rock slope’s safety and stability.

Key words: slope safety and stability, principal component analysis, combinatory weighting, fuzzy theory, comprehensive assessment