Abstract: The fluctuation of reservoir water level and rainfall are major inducing factors of reservoir landslide deformation, with deformation responses exhibiting significant time lag. To address the limitations of current reservoir landslide displacement prediction models in accounting for this time lag effect, a displacement prediction model based on VMD-SSA-LSSVM, incorporating time lag effects and effective rainfall, is proposed. Using the Baijiabao landslide in the Three Gorges Reservoir area as an example, monitoring data analysis and wavelet transform were applied to identify that the rapid decline of reservoir water level plays a dominant role in landslide deformation, while the influence of rainfall is relatively minor. Wavelet phase analysis and Granger causality tests were used to determine the lag times of reservoir water level fluctuation and rainfall on landslide deformation, and the effective rainfall in the preceding period was calculated. The displacement sequence was decomposed into trend, periodic, and random components using variational mode decomposition (VMD). The influence factor sequence, after incorporating lag time and effective rainfall, was further decomposed into high-frequency and low-frequency factors, and suitable factors were selected for each displacement component. Finally, the least squares support vector machine (LSSVM) model, optimized by the sparrow search algorithm (SSA), was used to predict and accumulate landslide displacements, and the predicted cumulative displacement was compared with other models. The results show that the proposed model achieves a goodness of fit of 0.9980 and a root mean square error of 1.9125 mm. Its prediction accuracy exceeds that of models that either ignore time lag or consider time lag without incorporating effective rainfall. This model provides a valuable reference for disaster prevention and mitigation of similar reservoir landslides.