Landscape connectedness, ecosystem service value (ESV), and resident welfare are intricately interconnected, and understanding their relationships is crucial for promoting regional sustainable development. Utilizing six stages of land use data from 1995, 2000, 2005, 2010, 2015, and 2020 in Xining City, Qinghai Province, this study identified landscape connectedness changes by landscape pattern indices, calculated ecosystem service value by the equivalent factor method, and quantified residents’ well-being by comprehensive evaluation indices. To investigate the coupled and coordinated states among the three, a coupling coordination model is adopted. Furthermore, the relative development degree model is employed to reveal the relative developmental level of the three, clarify the lagging factors in their coupling coordination. The gray relational model is employed to identify key factors affecting the coupling mechanism. Key findings include the following: (1) The development trend of landscape types was moving toward diversification and balance, and the total value of ecosystem services has been declining. The comprehensive level of resident welfare has increased annually, but the structure has changed. (2) The coupling relationship among landscape connectedness, ecosystem service value, and resident welfare is strong and has remained at a high correlation level but has been in a state of discord. (3) The main constraint of the discord in the early stage was resident welfare, but the constraining factors in the later stage shifted to ecosystem services and landscape connectedness. The largest patch index and water resource supply were the key influencing factors in the system coupling mechanism. The research findings can provide a reference for the sustainable development of Xining City, regional land use policies, and ecological intervention planning.