The problems with interviewing and selection of graduated students to work and occupy the right job according to their qualifications, still presents a great challenge for employment organization, and IT companies. There are several methods by which one can predict the appropriate job that is qualified to person's skills, but none of them is quite accurate. ANN models are very effective in predicting. In order to test the effectiveness of the developed ANN models, coefficient of determination and likelihood function that minimizes the root mean squared error were used. This research investigates how IT jobs predicting changes with respect to the skills and experiences, knowledge of the graduated students' situations by using back propagation artificial neural networks (ANN). To achieve the task result, data sets were taken from 50 graduated students. The data used in the multilayer feed-forward neural networks uses back-propagation algorithm models are arranged in a format of 35 input factors parameters for testing , and 35 input factors for training for 50 persons that cover IT job skills.