Abstract: The BP neural network and fuzzy comprehensive evaluation were used to predict slope stability and the 6 indexes, such as friction angle, cohesive force, severe, slope angle, pore pressure ratio and slope height were used as the evaluation factors. The stability of the slope was respective evaluated by using BP neural network and fuzzy comprehensive evaluation combining with the examples of 60 slopes. The results show that the accuracy of BP neural network is more accurate, and the fuzzy comprehensive evaluation can be used to achieve the classification of evaluation.