<p>Due to rapidly varying hydraulic boundary conditions, reservoir-induced slope instability remains difficult to monitor and forecast. This study aims to quantify the landslide reactivation mechanism at a reservoir shoreline in central Poland and to identify relevant hydrological thresholds for early warning. To achieve this, we integrate time-lapse electrical resistivity tomography (TL-ERT) with archival airborne photogrammetry, ALS LiDAR–derived digital elevation models (DEMs), and rainfall records, relating subsurface moisture dynamics to geomorphic change. The methods employed are operational: repeat TL-ERT sections on fixed profiles, coregistered multitemporal DEMs for elevation change, and standardized rainfall metrics to test threshold behavior. The results show the following: Recurrent reactivation is concentrated within zones of drawdown-controlled saturation contrasts. Spatially variable displacement fields are consistent with subsurface low-resistivity pathways. Quantitative hydrological thresholds combine short-term rainfall bursts with antecedent water-level states, markedly increasing reactivation probability. The inferred process sequence links reservoir level fluctuations, infiltration, and the formation of perched water to reduced shear strength along pre-existing structures. The implications include a hybrid early warning concept that combines simple, transferable rainfall/level thresholds with periodic TL-ERT checks to validate subsurface conditions where optical data are ambiguous. The novelty lies in coupling TL-ERT with historical airborne datasets to resolve the drivers and timing of reservoir-controlled reactivation, providing a practical template for long-term monitoring at similar shorelines.</p>