Soil infiltration rate is an important index in evaluating functions of soil and water conservation, which depends on many internal and external factors, and is closely correlated with forest types, soil structure, soil types, and intensity of rainfall and so on. It is also the a key factor in modeling soil water movement and erosion process where, The goal of this study was to determine the effects of reforestation methods that employ different water conservation forests on soil infiltration capability in northern water source area of Chongqing City. We used bare land as a control. We measured soil initial infiltration，stable infiltration，average infiltration and accumulative water quantity as soil infiltration capability, and determined the relationship between soil infiltration capability and soil physical and chemical properties under eight models of water conservation forest practices. The results showed that the soil infiltration capability values were significantly different in different forest models and different soil layers in the same forest. The soil infiltration capability declined with the increase of soil depth.Vegetation covers had no effect on the haplite soil layers. We found that the infiltration capability of all forest plots was higher than that of bare land. The order of soil infiltration capability from high to low was: Lindera kwangtungensis Cunninghamia lanceolata forest>Pinus massoniana Cryptomeria fortunei Hooibrenk forest>Gordonia acuminate Symplocos setchuanensis forest>Phyllostachys pubescens Symplocos setchuanensis Pinus massoniana forest>Pinus massoniana Lindera kwangtungensis forest>Pinus massoniana Gordonia acuminate forest>Lindera kwangtungensis Adinandra bockiana fores>Phyllostachys pubescens forest >bare land, and the test eight models of water conservation forest could be classified into four groups, based on soil infiltration capability. The first group, L. kwangtungensis C. lanceolata forest and P. massoniana C. f. Hooibrenk forest, had the strongest soil infiltration capability, and the second group with a stronger soil infiltration capability was G. acuminate S. setchuanensis forest. P. pubescens S. setchuanensis P. massoniana forest, P. massoniana L.kwangtungensis forest, L. kwangtungensis A. bockiana forest and P. massoniana G. acuminate forest were classified into the third group with a relatively strong soil infiltration capability, while P. pubescens forest and bare land belonged to the fourth group. In the eight reforestation models, L. kwangtungensis C. lanceolata forest, was best in improving soil infiltration capability. The role of P. pubescens forest, however, was the least. Because bamboo pure forest cover was largest in the study region, and the bamboo forests played a poor role in improving soil structure and infiltration capability, bamboo forests should be managed by increasing the proportion of other species of mixed forest soil to enhance soil infiltration capability. A common empirical infiltration model could give the best performance on the infiltrating processes in the eight forest models, followed by the Kostiakov and the Horton model. Correlation analysis showed that soil infiltration capability was greatly affected by soil physical and chemical properties. Based on the correlation analyses, nine factors, which have extremely significant or significant effect on soil infiltration capability, were selected. The comprehensive parameters of soil infiltration (α) and soil physical and chemical properties (β) were obtained by the Principal Component Analysis, and the linear regression models of five indices of soil infiltration capability and the comprehensive parameters (α and β) was built. These research results may be useful for local reforestation campaigns in term of tree species and forest community selections.