Abstract: Neural network technology is introduced in establishing a model to predict water quality of the Yellow River,and Levenberg-Marquardt(LM) algorithm is used to improve its prediction precision.This model is applied to prediction of water quality of the section from Xiaolangdi to Huayuankou of the river.Performance of the prediction demonstrates that the neural network model and the LM algorithm would be adopted with better results in establishing an input-response relationship of inflowing pollutants with water quality of the functional zone of the Yellow River.