Abstract: This paper uses the combining method of ANP and fuzzy cluster analysis in karstic collapse. It considers the factors such as mining activities, groundwater movement, karst development and cladding features. It is based on a great deal of practice. It further uses ANP to make a comprehensive analysis towards the influencing degree among various factors and adopts the method of Fuzzy clustering analysis to do clustering in dangerous areas. It is successfully applied to the prediction of karst collapse in Pb-Zn mining area in Gudan. The results indicate that carrying out mining activities is the mainest factors resulting in karst collapse. Other factors have effects on it as well. The study area can be divided into high risk area and low risk area. The results are in line with the empirical value and the measured value, which indicates that this method is reasonable and feasible, and provides a new method and thinking for the evaluation and prediction for dangerous area of karst.