Software engineering is incomplete without Software reliability prediction. For characterizing any software product quality quantitatively during phase of testing, the most important factor is software reliability assessment. Many software reliability growth models (SRGM) which is used for predicting in software reliability, however, no single model can give accurate prediction. For this the Artificial Neural Network (ANN) based software reliability model is introduced. In this thesis ANN based software reliability models for better reliability prediction in a real case is described and the growth of software reliability using ANN based model is presented. We proposed a neuro-genetic approach for the ANN based software reliability model by optimize the weights of the network by using proposed genetic algorithm (GA). Training the ANN using Backpropagation algorithm (BPA) to predict the software reliability is the first action. Than train our model global optimize the weight of the networks by using the proposed GA. Using two datasets contain cumulative executive time and cumulative no of software failures are applied to the proposed models. These datasets are obtained from software projects. Then it is observed that the results obtained indicate a significant improvement in performance by using genetic algorithm in ANN based software reliability models over the normal algorithm of ANN based software reliability models. Numerical and graphical explanations show that proposed model for software reliability prediction since its fitting and prediction error is much less relative to the normal algorithm of ANN based software reliability model.