[Objective] The spatial relationship between ecosystem service and residents’ well-being in a farming-pastoral zone was studied in order to determine the way in which ecosystem management could produce sustainable development of the social economy. [Methods] The spatial relationship and dynamic features of ecosystem service and residents’ well-being in fourteen counties (districts) in the Sanggan River basin (Shanxi section), a typical area of the farming-pastoral zone, in 2010, 2015, and 2020 were analyzed by using the entropy weight method and cluster analysis based on the modified equivalent factor method of ecosystem service value. [Results] ① From 2010 to 2020, both the per capita ecosystem service value and residents’ well-being in the Sanggan River basin (Shanxi section) showed an upward trend. The lower values of per capita ecosystem service occurred in the central and southwest part of the Sanggan River basin, while the high values of residents’ well-being were located in the southwest and northeast direction along the Sanggan River channel. ② The spatial correlation models of ecosystem service and residents’ well-being in the study area were: high service and high well-being, high service and low well-being, high service and medium well-being, low service and high well-being, low service and medium well-being, low service and low well-being, medium service and high well-being, medium service and medium well-being. From 2010 to 2020, the spatial correlation models of ecosystem services and residents’ well-being in the Sanggan River basin (Shanxi section) had changed mainly in the southwest and northeast direction. ③ The Sanggan River basin (Shanxi section) could improve regional ecosystem services and residents’ well-being by promoting the transformation and development of resource-based economy and strengthening the protection and restoration of land space ecology. [Conclusion] From 2010 to 2020, the value of ecosystem services and the well-being of residents in the Sanggan River basin have shown an increasing trend, and their spatial correlation patterns show obvious spatial differentiation.