In modern industry, metal pipelines are an important means of transporting energy. However, the pipe can corrode with time elapsed. Corrosion is one of the major causes of failure in gas pipelines. The purpose of this paper is to present a probabilistic approach to predicting the residual life of gas pipeline with corrosion defects. The proposed method consists of three steps. First, instead of deterministic one, a stochastic model is adopted to deal with the variability in corrosion growth rates. Second, the failure probability of the pipeline is obtained by using PCORRC, one of widely used codes to calculate the pipe failure pressure. Third, the remaining pipeline life is predicted by using ISO 16708, a global standard of limit state method on pipeline reliability. The feature of the presented method is that it accommodates the probabilistic behavior of corrosion growth and deals with the uncertainty of various pipe parameters. This makes it possible to obtain more realistic prediction results. An illustrative example is given by using an inline inspection data. Finally, based upon the residual life prediction, a future framework is provided for the pipeline prognostics management required in the era of Fourth Industrial Revolution.