Sediment deposition carried by flood flow is the main cause of reservoir sedimentation. This can be reduced by an appropriate operational strategy of flood flow and sediment in the reservoir. High-precision and large-scale hydrodynamic models to predict flood propagation and sediment transport in reservoirs are extremely important for an efficient flood forecasting and real-time joint regulation of water and sediment in reservoirs. In the present study, the three-dimensional (3D) numerical semi-implicit cross-scale hydroscience integrated system model (SCHISM) was adopted to model the flood propagation and sediment transport in the approximately 280-km-long reach in the Three Gorges Reservoir. This model is mainly focused on analysing the asynchronous movement characteristics of flood propagation and sediment transport and the operational strategy of sediment peak regulation. The flood event in July 2013 was reproduced by the numerical model, which was validated by a comparison with the measured data. The results indicated that the numerical model has the ability to accurately simulate the flood propagation and sediment transport processes. The time that the sediment peak lags behind the flow discharge peak increases as the flood waves propagate downstream, reaching 8.1 days at the dam site. During the rising period of the flood, the discharged flow is lowered to reduce the flood peak, and when the sediment peak reaches the dam, the discharged flow is increased to release high concentration sediment during the flood recession period so as to reduce sedimentation in the reservoir. The model results agreed well with the measured results. The 3D numerical model can be used for the real-time prediction of the arrival time of the flow discharge and sediment peaks for the joint regulation of water and sediment in the Three Gorges Reservoir.