Abstract: To precisely assess the ecosystem services supply-demand risk (ESSDR) in the Wuhan metropolitan area (WMA), safeguard ecological security, and support the construction of a low-carbon green metropolitan area, six typical ecosystem services were selected. By integrating the supply-demand competition-based two-step floating catchment area (2SFCA) model with service path attribute network theory, an ecosystem service spatial flow model was constructed. This model overcame the limitation of existing models in insufficiently accounting for supply-demand intensity, distance, topography, and other factors, enabling the objective grid-scale simulation of the entire "generation–flow–consumption" process of ecosystem service flows. In combination with the entropy weight method and ecosystem service demand framework, the spatiotemporal evolution of ESSDR in the WMA from 2002 to 2022 was quantitatively analyzed. The optimal parameter-based geographical detector was adopted to identify driving factors, with the best parameters determined by comparing the maximum 90th percentile of q-values across multiple scales to ensure result robustness. The results showed that the WMA's ESSDR was relatively severe and continued to deteriorate. Under the influence of the siphon effect of the core area, the overall deterioration gradient followed the order of Southern > Eastern > Core Area and its surrounding regions > Northeastern, and the average comprehensive ESSDR index decreased from −0.61 in 2002 to −1.93 in 2022. Spatially, ESSDR generally presented a "core-high, periphery-low" pattern. Population density was the primary driving factor of comprehensive supply-demand risk; over time, the secondary driving factor shifted from rainfall to construction land, reflecting the growing influence of socioeconomic factors. Furthermore, multiple factors exhibited synergistic effects, with particularly strong explanatory power when interacting with population density.