Under the long-term geological action, the saturated permeability coefficient of soil shows natural strong spatial variability. In order to explore the influence of saturated permeability coefficient variation on the failure probability of slope, the Green-Ampt model was improved and the relationship among the depth of wetting front, the safety factor of slope and the duration of rainfall was determined. Then, Monte Carlo method was used to establish the analysis framework of variation of saturated permeability coefficient, and the  influence of saturated permeability coefficient variation on the failure probability of slope under different rainfall intensity was discussed. The results show that in the initial stage of rainfall, the smaller of the coefficient of variation of saturated permeability is, the smaller the failure probability of the slope is. With the increase of rainfall duration, the failure probability of the slope increases gradually, and with the increase of the coefficient of variation of saturated permeability, the failure probability of the slope becomes smaller. In the critical rainfall range, the failure probability of slope instability rises suddenly, and the critical rainfall range becomes concentrated with the increase of rainfall intensity.