< p >预测treatme污染物去除的nt plant of Shahid Salimi industrial town is performed in this study using artificial neural network. The required data of this treatment plant are achieved by 162 records after eliminating the repeated and uncompleted data. The appropriate inputs (BOD, COD, and TSS) are chosen by the correlation analysis in terms of having the highest correlation with the output parameters of the treatment plant, considering the small size of data sets and the need to simplify the model. The architecture of the network is used to make the proper prediction, which uses one neural network to predict all the output parameters (BOD, COD, and TSS). This network with 20 neurons in two hidden layers could predict the output of the treatment plant with good accuracy. Excellent results indicating the success of the modeling are obtained using the mentioned architecture.