Abstract: The ENSO（El Niño-Southern Oscillation）cycles have a significant impact on runoff evolution in the Yangtze River basin, highlighting the importance of analyzing runoff patterns and teleconnection mechanisms. This study uses a Non-homogeneous Hidden Markov Model (NHMM) to dynamically simulate sea surface temperature anomalies (SSTA) in the Pacific Ocean, identifying five distinct ENSO states and analyzing the associated runoff, precipitation, and circulation data. The results show that during strong El Niño events, moisture convergence and anticyclonic conditions in Eastern China lead to substantial moisture transport to the middle and lower reaches of the Yangtze River, causing anomalous rainfall and increased runoff. In contrast, during strong La Niña events, the southwest and northwest regions of the Yangtze River basin experience anomalous rainfall due to a significant moisture belt and cyclonic activity around 10°N. The vertical atmospheric velocity (Omega) and geopotential height (GPH) align with water vapor transport. This study enhances the accuracy of runoff forecasting and provides technical support for disaster prevention and mitigation.