Abstract: In the context of climate change and urban expansion, it is of great significance to conduct in-depth research on the spatiotemporal evolution and influencing factors of ecosystem services. Since the key parameters of the InVEST model are highly subjective and cannot be dynamically changed across different regions, making it difficult to reflect the spatiotemporal heterogeneity of real ecological processes. This study takes five major urban clusters in the Yangtze River Basin as the research object, utilizing the ecological significance and frequency distribution of remote sensing ecological indices, the habitat quality parameters of the InVEST model are dynamically optimized. Soil conservation, water conservation, and habitat quality in the Yangtze River Basin urban clusters from 2000 to 2020 are calculated, and the ecosystem service impact factors over the past two decades are analyzed using random forest and SHAP algorithms. The results show that: (1) The ecosystem services of the urban clusters in the Yangtze River Basin show significant upstream and downstream gradient characteristics. The upstream has better ecological maintenance, the midstream fluctuates significantly and is more sensitive to human activities, and the downstream shows obvious ecological vulnerability. (2) Hydrology, climate and vegetation cover are the core foundation for the improvement of the three types of ecosystem services, playing a positive role in water conservation, soil conservation and habitat quality; pollutants and human activities are the main negative driving factors, especially in the middle and lower reaches, which seriously lead to ecological degradation. (3) The spatiotemporal evolution of ecosystem services is jointly determined by the coupling relationships of various factors. Climate alone has a weak impact, but pollution, land use intensity, and human disturbance brought about by urban expansion directly affect hydrothermal conditions and vegetation distribution. Studies have found that, under the background of urbanization, the urban clusters in the upper reaches of the Yangtze River Basin are ecologically stable and dominated by natural processes, while the urban clusters in the middle and lower reaches are severely affected by human activities. The overall ecological difference pattern of the basin is jointly shaped by ecological processes and human disturbance.