The complex topography makes the small watershed have many ecosystem services, such as production, water supply and regulation, and significantly regional differences. However, the balance of ecosystem service and the integration measurement of external driving mechanism in the small watershed, which have always been a difficult problem in the research of geography, ecology and economics. This paper takes Qihe River Basin in the south of Taihang Mountain as the research object, using CASA and InVEST models estimate carbon sequestration, soil conservation and water resources supply services respectively, based on the multi-source data such as meteorological data, land use/cover data, Normalized Difference Vegetative Index(NDVI). The spatial overlay analysis method and correlation analysis are used to measure the ecosystem services trade-off relationship, and the spatial impacts of natural environment and human activities on the ecosystem service trade-off relationship are further analyzed by factor detection and interaction detection in the geographical detector. The results indicate that ① from 2000 to 2015, the average annual carbon sequestration is 375.55 gC m^-2 a^-1, and the annual deceleration rate is 18.20 gC m^-2 a^-1 in the Qi River Basin; the annual average soil conservation is 396.72 t hm^-2a^-1, and the annual decline rate is 1.2 t hm^-2 a^-1; the average annual water resources supply is 67.26 mm/a with decreasing rate of 0.16 mm/a. ② The ecosystem service relationship is dominated by strong balance and low coordination. There are significant spatial differentiation characteristics during the research period, low coordination distribution area reduced 16.21% and strong balance increased 11.85%, indicating the ecosystem service capacity of the river basin is increasing;In the pairwise relationship of ecosystem services, carbon sequestration and water resources supply, water resources supply and soil conservation show a trade-off relationship, while carbon sequestration and soil conservation show a synergistic trend.③ Geodetector result show that NDVI (0.306) > temperature (0.241) > land use degree (0.002) in terms of the carbon sequestration and soil conservation; for carbon sequestration and water supply, vegetation NDVI has the strongest explanatory power (0.381); for water resources supply and soil conservation, the explanatory power of precipitation, temperature and elevation factors are higher than 0.5; the explanatory power of each influencing factor is obviously enhanced after interactive detection.