Accurate assessment of waterlogging susceptibility is important for maintaining productivity and ecological stability in oil palm plantations, particularly in the face of increasing climate-driven rainfall variability. Direct field measurements of plant and soil water status provide reliable information but are labor-intensive and impractical for repeated monitoring at plantation scale. This study proposes a data-driven predictive modelling framework to assess waterlogging susceptibility in oil palm plantations using UAV-derived multispectral information. Field and aerial data were collected at the Oil Palm Teaching Farm, IPB University, Indonesia, during four seasonal observation periods (July 2024, October 2024, January 2025, and April 2025). Multispectral orthomosaics were generated, and twelve vegetation indices were extracted at the individual palm canopy level. Soil moisture, leaf water content, and leaf greenness were measured in the field and used as reference data to develop predictive models based on Random Forest Regression, Partial Least Squares Regression, and Support Vector Regression. The modelling results showed that Partial Least Squares Regression provided the best performance for soil moisture estimation (R² = 0.61), while Random Forest Regression achieved high accuracy in predicting leaf greenness (R² = 0.83). In contrast, all models exhibited limited performance in estimating leaf water content (R² < 0.50), indicating low sensitivity of multispectral vegetation indices to variations in oil palm leaf water status. Consequently, waterlogging susceptibility mapping was conducted based on the integrated spatial patterns of predicted soil moisture and leaf greenness. The resulting susceptibility maps successfully identified palms vulnerable to prolonged waterlogging and associated productivity decline. These findings demonstrate that data-driven predictive modelling using UAV multispectral data can provide a practical and scalable approach for spatially explicit assessment of waterlogging susceptibility in oil palm plantations, supporting informed decision-making in precision and environmentally sustainable plantation management.