Fuzzy stochastic theory is applied to the slope reliability evaluation. Considering the fuzzy and stochastic features of random variables, the subordinate function of stability factor is introduced to perform fuzzy reliability analysis for the actual working condition of the slope. The secondary structure of fuzzy random limit state equation is proposed on the basis of combining the fuzzy stochastic theory with the response surface method. The response surface method is used on a structure firstly to obtain the reliability response surface function, based on which the complementary function of subordinate function is used as an auxiliary to carry on the secondary limit state equation to calculate the random fuzzy reliability of the slope. Rather than directly solving the fuzzy probability density function by integral calculus, which is complicated and time-consuming, this method is proved to be simple with fast convergence and high precision of calculation, thereby can be used widely. The slope stability can be reflected more accurately by the calculated reliability indexes. The method can be regarded as a new approach of studying the reliability of slopes.