Abstract: Landslide damming events occur frequently in the upper reaches of the Jinsha River, significantly affecting the evolution of the river network and the operation of cascade hydropower plants. To reveal the impact of an extremely large-scale landslide on the morphological characteristics of the mainstream and tributaries of the Jinsha River, this study employed multi-source satellite remote sensing data to identify the landslide extent from 2018 to 2025 at the Baige landslide site, as well as the morphology of the river systems before and after the landslide. Indicators such as river network density, sinuosity, and box-counting dimension were used to quantify the morphological changes in bank slope tributaries within the landslide area and the mainstream of the Jinsha River before and after the landslide. The results indicate that prior to the landslide, the bank slope tributaries exhibited a comb-like distribution, independently flowing into the mainstream. After the landslide, catchment conditions of slopes underwent significant changes, with an increase in slope runoff that hindered the formation of a complete slope drainage system. The average river network density on the left and right banks decreased by 2.7% and 29.6%, respectively, transforming the morphology into a dendritic pattern with two nearly interconnecting ends. The influences of the Baige landslide on the morphology of the mainstream were primarily concentrated in the landslide deposit segment and its upstream section, changing from approximately C-shaped to approximately S-shaped. The increase in riverbed height mainly affected the morphology of the mainstream during the dry season, with minimal impact during the wet season. After the landslide, sinuosity increased from 1.16 to 1.17. Three years after the landslide, the overall morphology of the mainstream tended to be stable. A single extremely large-scale landslide can easily trigger multiple phases of the "landslide-dammed lake-dam failure" disaster chain. Therefore, there is an urgent need to strengthen research on real-time monitoring, early warning and prediction technologies throughout the entire process.