Abstract: A river quality management model is proposed for the optimal waste load allocation in a tidal river basin,taking into account the impacts of stochastic pollutant discharge and dynamic hydrological conditions on water quality during a tidal cycle.The model is known as the maximization of total allowable waste load,subject to meeting water quality standards at checkpoints along the river.The waste load is regarded as a stochastic variable following the log-normal probability distribution based on statistical data,and the constrains on water quality levels are expressed in a probability form.A genetic algorithm is used to solve this stochastic programming model.A case study shows that the optimal waste load allocation can achieve water quality standards in stochastic conditions,and the model's outcomes are clearly different from those of the conventional deterministic linear programming model 1 The computational results imply that a genetic algorithm is effective to solve complicated stochastic programming models.