Abstract: Traditional analysis methods for reliability evaluation of corroded pipelines only focus on certainty factors, but without considering the effect of uncertainty factors. Limit state function of corroded pipeline failure is built on the basis of fundamental theories of reliability analysis methods. The Monte Carlo method is adopted to analyze comprehensively the effects of uncertainty factors (e.g. pipeline, pipeline defect size, material natures, and load) on pipeline reliability. Sensitivity analysis is conducted on each stochastic parameter of corroded pipelines, and it is revealed that corrosion depth and radial corrosion rate are two key factors affecting reliability of pipelines. The effect of different numerical characteristics values of corrosion depth and radial corrosion rate on pipeline reliability is investigated by using MATLAB to simulate the corrosion development process of oil and gas pipelines. The life of the pipelines is also predicted. Obviously, the results of pipeline reliability analysis play an important role for pipeline maintenance decision.