[Objective] To construct a comprehensive multi-temporal landslide inventory across the Eastern Himalayan Syntaxis and quantify landslide-driven erosion rates, thereby revealing the geomorphological significance of landslide processes in this region. [Methods] The study employs the Particle Swarm Optimization (PSO) algorithm for change detection of the Normalized Difference Vegetation Index (NDVI) from remote sensing images, enabling the construction of a multi-temporal landslide inventory for the Eastern Syntaxis from 1987 to 2021. The landslide erosion rate was calculated using an empirical landslide area-volume relationship. Additionally, the study explores the factors inducing landslide processes by considering climatic and topographic parameters. [Results] A total of 1 323 landslides were identified in the study area between 1987 and 2021, with the highest occurrence of 389 landslides recorded between 2017 and 2021. The landslides predominantly occurred on both sides of the river valleys near the Yarlung Tsangpo River's Great Bend. The landslide erosion rates in the study area ranged from 0 to 76.06 mm/a, with an average rate of 0.44 mm/a. These rates show a decreasing trend from the Great Bend section of the Yarlung Tsangpo River outward. The erosion rates are comparable to the exhumation rates of geological-scale rock bodies and the millennial-scale basin-wide average erosion rates. Landslide occurrences are associated with rainfall events and seismic activities, primarily developing on south-facing slopes and clustering within an elevation range of 1,500 to 3,000 meters and slopes of 35o to 45o. [Conclusion] Landslides represent the dominant erosion process in the Eastern Himalayan Syntaxis. Rainfall, influenced by the windward slope effect, concentrates on south-facing slopes, driving the concentrated distribution of landslides on these slopes. Moreover, precipitation also triggers landslides by enhancing river incision, which steepens the adjacent slopes.