A stability analysis model based on modified Morgenstern-Price method is established from both perspectives of design and check in order to assess the stability of pile stabilized slope, and the expressions of sliding thrust for stabilizing pile and factor of safety for slope are derived. In this model, an adaptive genetic algorithm is introduced to form the optimization model of slope stability analysis. The algorithm is used to search for the critical failure surface of pile stabilized slope. Then, the influence of the installing position of stabilizing pile on the slope stability is investigated under two situations: obtaining the sliding thrust as given factor of safety and determining the minimum factor of safety as given sliding thrust respectively. The results indicate that this approach can not only seek out the critical failure surface with non-circular shape which is more suitable for the reality, but also determine either the sliding thrust of stabilizing pile or the minimum safety of factor for slope as well. It can be concluded that under the same circumstances, stabilizing piles should be supposed to be installed in the middle part of the slope, where the reinforcement effect of stabilizing piles can be maximized.