In order to study the difference of soil infiltration performance and its influencing factors in different slopes of typical vegetation in Loess Alpine region, the infiltration process of different vegetation types was measured by constant head method. Soil porosity, soil particle-size distribution, organic matter and water-stable aggregates were also analyzed. The results showed that:(1) there were significant differences in infiltration rates among different vegetation types. The initial and stable infiltration rates changed in the order of Picea crassifolia > Sabina przewalskii > Larix principis-rupprechii > barren grassland. The soil infiltration rate increased with the increase of soil porosity, aggregate content and the decrease of bulk density along the slope. (2) The fitting effect of Horton was the best on different slope positions and vegetation types, and the general accuracy for the lower slope is better than that for the upper slope, while the general empirical model and Kostiakov model are poor in fitting effect. (3) Soil infiltration capability was significantly correlated with soil porosity, organic matter and water-stable aggregates, among which the main factors affecting the stable infiltration rate were bulk density, > 0.25 mm aggregate, and total porosity. The main factors affecting the initial permeability rate were > 0.25 mm aggregate, total porosity and capillary porosity. Path analysis shows that total porosity is the main decision variable affecting the initial infiltration rate, while capillary porosity is a limiting factor; >0.25 mm water-stable aggregate mass fraction is the main decision factor affecting the steady infiltration rate. This paper provides a scientific basis for soil infiltration simulation and vegetation allocation in Loess Alpine region.