Abstract: The middle reaches of Changjiang River face challenges in basin flow regulation due to its complex hydrological processes and frequent extreme climate events.Therefore, analyzing the dynamic flow coupling relationships among key hydrological sections and constructing runoff prediction models were crucial for revealing the basin′s hydrological response mechanisms and optimizing water resource allocation.This study utilizes daily-scale hydrometeorological data from five key hydrological cross-sections in the middle reaches of Changjiang River (Yichang, Shashi, Jianli, Luoshan, and Hankou).By integrating upstream inflow, regional precipitation, temperature, and historical flow records at these sections, an Autoregressive Distributed Lag (ARDL) model was constructed to quantitatively analyze inter-section flow lag response characteristics and predict runoff.Results indicated that upstream inflow, regional precipitation, and temperature exhibit lag periods ranging from 1 to 5 days in influencing daily discharge at downstream cross-sections, with lag factors varying across different sections.The ARDL model performed well in predicting the flow volume of all five sections.During the validation period, its coefficient of determination R² exceeded 99%, and the mean square percentage error (MSPE) was below 5%.The ARDL model can provide a reference for runoff forecasting in the middle reaches of Changjiang River.