Clarifying ecosystem service interactions under urbanization facilitates integrated ecological governance. This study took the rapidly urbanizing Su-Xi-Chang agglomeration in the Yangtze River Delta as the research area and applied the InVEST model to evaluate spatiotemporal variations in carbon storage, soil retention, water yield, and habitat quality from 2000 to 2023. The coupling coordination degree model and self-organizing map model were adopted to analyze urbanization–ecosystem interactions and the evolution of ecosystem service bundles, while the GeoDetector model was used to identify key socio-ecological drivers. The results reveal significant spatiotemporal divergences among the four ecosystem services. Carbon storage and habitat quality declined, while water yield and soil retention increased. Carbon storage and soil retention hotspots were persistently concentrated in the Yixing–Liyang Mountainous Area. Pairwise correlations further show that the synergy between water yield and soil retention strengthened over time, whereas their trade-offs with habitat quality intensified. The coupling coordination between urbanization and ecosystem services remained at a primary level yet exhibited a clear trajectory from imbalance toward optimization after 2010. Four ecosystem service bundles were identified, among which the urban-function-dominated bundle expanded by 190.39% from 2000 to 2023, indicating progressive spatial lock-in of urban land use. Differentiated zoning and targeted restoration strategies are proposed to reconcile urban development and ecological conservation. This study offers scientific references for ecological spatial optimization and refined ecological management in the Yangtze River Delta.