Abstract: Considering the high nonlinearity and uncertainty of influence factors in a slope system and making full use of the ad-vantages of the rough set theory and genetic programming (RS-GP model), a novel method based on the influencing factors of slope stability was brought up to calculate the safety factor of a slope. This method can reduce the structure scale of a genetic programming prediction model by analyzing the data characteristics of learning samples, calculating the significance of attributes and reducing rules. The results of an ANN model and the RS-GP model show that the proposed method has such merits as fast computing speed, high fault-tolerance capacity and high precision.