Only the spatial variability of soil saturated hydraulic conductivity was considered, while that of shear strength parameters was ignored for investigating rainfall-induced landslide mechanism and reliability of infinite slopes. This paper proposes a modified Green-Ampt model to determine the distribution of soil moisture content and the wetting front depth for slope reliability analysis under different rainfall durations. Rainfall-induced landslide mechanism considering the interaction of the spatial variability of multiple soil parameters and the rainfall infiltration is explored. The probabilities of slope failure under different rainfall durations are compared. A small coefficient of variation of saturated hydraulic conductivity generally induces a smaller probability of slope failure during the initial period of rainfall, but it will result in a large probability of slope failure as the rainfall sufficiently progresses. Additionally, the uncertainty of slip surface location due to the spatial variability of shear strength parameters will greatly affect the slope stability during the initial period of rainfall. As a result, the slope fails along the weak zones induced by the spatial variability of shear strength parameters besides along the wet fronts and impermeable layers, the larger probability of slope failure is produced. As the rainfall progresses, the advance of the wet front will become the crucial factor that affects the slope stability wherein the slope will fail mainly along the wet fronts. Consequently, the probabilities of slope failure are similar whether or not the spatial variability of shear strength parameters is taken into account.