Many reservoirs and associated downstream ecosystems located in the Asian monsoon climate region are under increased pressure from the long-term negative effects of turbid flood runoff. Despite the ubiquitous use of turbidity (CT) as a barometer of water quality and environmental “health”, CT modelling studies have been rare due to lack of detailed experimental data required for validation. This study explored the fate and transport of a turbid density flow entering a stratified reservoir (Daecheong Reservoir, Korea) through application of a coupled three-dimensional (3D) hydrodynamic and particle dynamics model and subsequent validation of the predictions against an extensive data-set collected during the flood season of 2004. The turbidity model simulated multiple size groups of suspended sediment (SS) and used site-specific SS–CT relationships to convert between field measurements (CT) and model state variables (SS). The model showed good performance in reproducing the reservoir thermal structure, flood propagation dynamics and the magnitude and distribution of turbidity in the reservoir. The turbidity modelling framework developed in this study can be adopted within a real-time management system for forecasting of turbidity and to support adaptive control of withdrawal facilities in reservoirs that suffer from turbid flood runoff.