In flow performance relationships (IPRs) represent the relation between well flow and the bottom hole pressure; the performance prediction of IPRs plays a fundamental role in the optimization of production operations and subsequently to design an artificial dislocation to select the most appropriate production scenario. The purpose of this study is to consider the effect of several important parameters on IPR curves by the utilization of a multilayered perceptron artificial neural network. The inputs of this system are the well pressure, the average reservoir pressure, and the maximum flow rate, while the output of this system was assumed pumped oil. Since then by using an artificial neural network code with MATLAB software and coupling it with the genetic codes, a random search algorithm to predict the IPR curves are plotted. Consequently, according to the results of this study, the average reservoir pressure has a profound effect on the IPR curves and the genetic algorithm plays an essential role in determining the weights, bias, and the structure of the artificial neural network. in respect of the way, four percent error in actual data from the network in the test data section indicates that this algorithm is based on a random search can well identify the relationship between inputs and outputs. Furthermore, artificial neural networks can be used with good precision and credibility rather than a reservoir simulator to generate other data due to the acceptable results of data matching.