Abstract：Fuzziness exists in thinking and judgment of human brains during the procedures of analysis and design of slope engineering，which will bring fussy indefiniteness to some extent to the whole courses of the analysis and design. Thus slope engineering，in fact，is a dynamic，fuzzy，open and complicated nonlinear system，which makes it difficult to evaluate complicated slopes conforming to reality by the traditional analysis method. Artificial neural networks (ANN) and fuzzy reasoning method，which have been applied to the evaluation of slope stability respectively，have shortcomings of their own so that ANN were used to constitute T-S¢s fuzzy reasoning system and the ANNs were trained by hybrid genetic algorithms (HGA). A new fuzzy reasoning system driven by HGA-based ANN，which was used for estimation of slope stability，was constructed. Based on the 80 slope cases that are collected from the worldwide practical slopes，a new estimation method for slope stability is built up. The proposed HGA-ANN-driven fuzzy reasoning system is actually a weighted combination model. Compared with the maximum likelihood method and BP neural network，the presented model has higher predicting accuracy.
