A total of 69 sites were deployed in the main and tributary streams of the Yangtze River Basin to evaluate and analyze the spatial autocorrelation of main water quality parameters and to explore the influence of multiple environmental elements on the water quality of the Yangtze River at different spatial scales. The results showed that nearly half of the sampling sites had water quality of category 3 or above， with obvious differences between the main streams and tributaries. DO， TP and TN had obvious aggregation trends on the whole， but had specific distribution characteristics in the local space. In addition， the relationship between water quality parameters and influencing factors at different spatial scales is complex， and the application of Geographically Weighted Regression model is always better than Ordinary Least Squares model. At the global scale， precipitation and water body area positively contributed to water quality and were the main factors influencing DO and NH₃-N， while temperature and nighttime lighting had negative effects and were the key correlates of TP， TN， permanganate index， and WQI （composite pollution status index）， with temperature also significantly influencing water body pH. At the local scale， there were complex correlations between multiple influencing factors and water quality parameters， with differences in their roles at different spatial scales， mainly in the effects of NDVI （Normalized difference vegetation index）， population density and built-up land. In summary， this study thoroughly explored the main driving mechanisms of water quality changes at multiple spatial scales， providing an important basis for the protection and regulation of water quality in the Yangtze River.