Software reliability is the important attribute for complex computing systems to provide reliability could cause series issues such as extra cost, development delay and image of the software solution providers. Hence, ensuring the reliability of software before deliver to the customer is essential part for the company. Finding the error in right time with reasonable degree of accuracy helps to prevent the consequences. Several software reliability growth models developed and used to measure the trustworthiness based on development and testing phases with unrealistic assumption over the environment and applied Block box methodologies while constructing model. This paper presents well established statistical time series (S)ARIMA approach for developing a forecasting model that able to provide significantly improved reliability prediction. Using real time publicly available software failure sets, the prediction of proposed model is developed and compared with previously available reliability models.