Fast and accurate estimation in Peak flow is one of the major issues in water resources management that have basic role in the design of hydraulic structures and biological activities in basins. So that a proper assessment has a basic role in the success of administrative tasks. In this paper, using artificial intelligence methods (Multi-layer Perceptron Neural Network and The mixture of Multi-layer Perceptron Neural Network with genetic algorithm) to estimate yalfan river, s peak discharge in Yalfan,s sediment and hydrometer local station. For these two models, 8 variables have been considered as the inputs that includes Flooding rain,5 days rainfall that occurs before of the flooding day, cure number of the basin(CN) and basic discharge and finally peak discharge is consider as the output. In the artificial intelligence after preprocessing of the data, the optimal structure of the model is determined with input and output data, Evaluation Criteria and trial and error. In the final the mixture of artificial network and genetic algorithm that the neural network has been a input role, have been a good performance in runoff forecasting in Yalfan Basin.