Efficient management of credit risk in the banking system of Iran requires the implementation of a customer validation system. The high volume of banks' Non-Performing Loans (NPL) is one of the largest diseases in the banking system of Iran. This is the result of the lack of usage of data-oriented mechanisms in customer validation processes. Therefore, presenting such tools and mechanisms for validating and ranking of customers of the bank is very necessary. The main purpose of the present research is to provide a model for predicting the credit behavior of customers in a private bank in Iran, using the artificial neural network. Neural networks, while having higher precision, less volume of computing, and also the lack of need for some limiting assumptions, has more advantageous than other classical prediction methods. A feedforward neural network has been developed and its functional results have been compared with the network with the feed-forward neural network. Analytical results show that the proposed model has the ability to predict test data appropriately and has been identified with a lower error rate than the other model. So, this model can be the appropriate model for predicting customer credit behavior. Using this model makes it possible to classify the customers of the bank in two groups based on risky behavior with high precision.