Over the last several decades, researchers have studied the hydrological disciplines through observed data. But with the impacts of human activities, the observed data cannot show the real hydrological situation gradually. The quantificational analysis of hydrological response to human activities has become the indispensable foundation when reservoir, water control project and large scale farmland were built. This analysis is facing many unresolved problems, and it becomes the advance front of the field of hydrology and water resource. The effects of human activities on hydrology can be divided into direct effects and indirect effects. The direct effects are classified to irrigate farmland and forests, to build reservoirs and to carry out water control projects and so on. These activities change the quality, quantity and distribution of water resource directly. The indirect effects are urbanization, tree planting and agriculture and so on. These activities influence the water resource indirectly by changing the land use types and weather. In this study, the year of 1986-1987 and 2005-2006 were chosen as two study periods. Two land use maps on the base of two TM images of 1987 and 2005 were classified manually. A SWAT model was setup and carried out in Guishui River basin, Beijing China to simulate the hydrological response to human activities. The results indicated that, human activities had remarkable effects on the runoff. In the study periods, the direct effects influenced the runoff much more than the indirect effects did. The indirect effects increased the runoff, especially dramatically in flood season (June to August) with the variation percentage was 34.67%. The swelling stage (April to August) and falling stage (September to December) got faster and the runoff showed high relativity to the precipitation. Along with the increase of the precipitation, the relativity was also enhancing, and the upward non-uniform coefficient indicated that the runoff varied more. The direct effects decreased the runoff coefficient by 87%, and more obvious in flood season when the peak values almost disappeared. The decreased non-uniform coefficient indicated that the runoff became well-distributed. So the indirect effects brought more droughts and floods while the direct effects can improve the water operating factor and make fewer droughts and floods. In this paper,conclusions are: (1) In this study, SWAT model simulated the runoff of Guishui River basin reasonable. This study also enriched the experience of the applications of SWAT model in China. (2) The expansion of the urban occupied more and more farmland, and then the farmland encroached on forests and grass. The increasing impervious surface changed the ability of water conservation of Guishui River basin. (3) The direct effects and indirect effects had different mechanism, and the direct effects play the most important role in the changes of the runoff. Hence need to note that the direct effects should be never ignored or be confused with the indirect effects. And more exploration need to do to increase the depth of it. Only in this way can we better understand the hydrological response to human activities clearly and correctly.