Abstract: The optimizing sediment regulation will safeguard long-term reservoir operations and maximize the reservoir effectiveness. In this study, an optimization model for reservoir sediment regulation is developed based on the principle of efficiently removing sediment. The reservoir sedimentation is calculated using a one-dimensional water-sediment transport model, the back-propagation neural network is adopted in the model training processes, and a genetic algorithm is used to ensure the high performance in the model calibration. The model is applied to the Three Gorges Reservoir. The result shows that the model is capable of finding the balanced optimized solution for the reservoir sediment regulation and maximizing the reservoir power generation.