Abstract: Reservoir bank landslides, triggered by multiple factors, exhibit significant nonlinearity and spatiotemporal heterogeneity in their catastrophic evolution and dynamic responses. Effective hazard mitigation requires elucidating the influence of coupled rainfall infiltration and rapid reservoir drawdown under extreme conditions on landslide evolution mechanisms and dynamic response characteristics. This study employs the Material Point Method (MPM) to construct a two-dimensional numerical model of the Outang landslide (Three Gorges Reservoir Area), simulating its initiation and acceleration under coupled rainfall and reservoir drawdown conditions. By analyzing deformation and stability across different landslide sections under varying hydraulic scenarios, the evolutionary characteristics and failure mechanisms were identified. The results show that: (1) The stability of the Outang landslide is jointly controlled by rainfall and reservoir water-level changes. Drawdown primarily affects the primary sliding mass at the toe, while rainfall significantly impacts the stability of the tertiary sliding mass at the crest, consistent with field monitoring data; (2) Under combined rainfall and drawdown conditions, localized collapse occurs at the toe and overall sliding is observed at the crest, with no significant mid-slope instability; (3) Under extreme rapid drawdown and intense rainfall, significant displacement is confined to the toe and crest, ruling out large-scale translational sliding along the bedrock interface; (4) During failure initiation, a significant spatial discrepancy exists between the distributions of initial strain and initial displacement along the main sliding direction, a critical consideration for determining monitoring point placement. Utilizing large-deformation numerical simulation, this study investigates dominant factors controlling the long-term stability of giant paleolandslides along reservoir banks, providing a theoretical basis for early warning and mitigation strategies for analogous geohazards.