[Objective] To investigate the soil moisture response to rainfall in loess ground with different slopes and vegetation types. [Methods] Eight in-situ observation plots were established, including flat surfaces with bare soil, Artemisia cover, and Robinia pseudoacacia cover, 15° slope with bare soil, Artemisia cover, R. pseudoacacia cover, and R. pseudoacacia-Artemisia complex cover, and 35° slope with R. pseudoacacia-Artemisia complex cover. Soil moisture content and basic physical parameters of the profiles were measured. The wetting front depth after rainfall was observed and the impact of vegetation type and slope gradient on rainfall infiltration was analyzed. Additionally, laboratory model tests with six different slope gradients were conducted to study the relationships between slope gradient, as well as vertical infiltration depth, and infiltration thickness normal to the slope surface. [Results] The results suggest that vegetation types significantly affected the depth of the wetting front, soil moisture content, and saturation of slope soils. The R. pseudoacacia-Artemisia complex cover had the shallowest wetting front depth with a saturation of 73.0%. The bare soil exhibited the deepest wetting front and the lowest saturation, which was only 52.4%. The slope gradient had a minimal impact on the wetting front depth, soil moisture content and saturation for slopes with identical vegetation. However, after a number of short-term rainfall events, the moisture content and saturation in the wetting zone were slightly higher in flat ground than those in slopes. Laboratory model tests revealed that the vertical infiltration depth in slope was independent of slope gradient. In contrast, the infiltration thickness normal to the slope surface decreased with increasing slope gradient. [Conclusion] The study can provide a theoretical support for shallow landslide mitigation and soil erosion control for loess slopes.