Abstract: The geological uncertainty is often ignored in slope reliability-based design, even though the spatial variability of soil parameters is considered. A full probabilistic design method is proposed for the slopes considering the geological uncertainty and spatial variability of soil parameters. In the full probabilistic design framework, a generalized coupled Markov chain model is combined with a random field model to simultaneously characterize the geological uncertainty and spatial variability of soil parameters. The procedure for this method is presented. A slope is taken as an example for the reliability-based design using the borehole data in Perth, Australia. In order to illustrate the importance of considering the geological uncertainty and spatial variability of soil parameters in the slope reliability-based design, the reliability design results associated with two cases, i. e. only considering the spatial variability of soil parameters and considering both types of uncertainties, are compared. The results indicate that the proposed method can effectively conduct the slope reliability-based design considering these two types of uncertainties. If only the spatial variability of soil parameters is considered, the reliability design results mainly depend on the used geological profiles. If the geotechnical practitioners infer a geological profile with a higher proportion of strong soil materials than the reality, the resulting optimal design scheme will lead to dangerous slope. In the opposite case, the resulting