Historically, rivers as a major source to supply drinking water and agriculture are considered in human societies as a very effective in the formation of civilizations. At present, most of the northern rivers that are under the influence of human, cause a variety of changes and destruction of ecosystem. Pollution and human interfaces could be the most important factors to pollution caused by waste water of industrial, urban and rural, pollution resulting from disposal of pesticides used in agriculture, destruction of vegetation, construction of dam, and barriers under bridge, blocking the mouth of the river, illegal fishing. Hence, considering the importance of Talar River for supply agriculture water, and also pour pollution to in, to identify and assess the river water quality and provide the necessary relationship to estimate pollution and water quality. In current study, 72 samples were used during the years of 1391 to 1392 from 6 stations namely, weresk, pole sefid, Shirgah, Talar, Kiakola and Arabkhil. Then, NSFWQI index was estimated. Finally, with applying gene-expression programming and artificial neural network, the obtained models for determination of relation between water quality parameters and water quality index are used with high accuracy. To evaluate the performance of models, statistical parameter such as, root mean square error, mean absolute error, scatter index and correlation coefficient were used. Results showed that the proper functioning of artificial neural networks and planning methods in estimating gene expression index is NSFWQI. The obtained results revealed appropriate performance of that artificial neural and geneexpression programming to estimate NSFWQI index.