Abstract: In this paper, in order to predict the reliability growth of maintainable gas pipeline system, a mathematical model is established by combing the generalized gamma distribution and statistical theory. In this model, the digital simulation technology is adopted to predict the reliability growth of maintainable gas pipeline system and analyze the prediction precision with point estimation, interval estimation, standard deviation and proportionality coefficient. Moreover, the effectiveness of this model is verified by practical application. The study shows that the concept of life analysis for failure distribution can be used for reliability growth prediction of maintainable gas pipeline system, so as to effectively improve the prediction precision. Increased amount of data is conducive to avoid low precision of reliability growth prediction caused by insufficient data of small samples like pipeline. The model can provide theoretical guidance for pre-maintenance and maintenance.