The work aims to predict surface corrosion rate of metallic gas pipeline. Surface corrosion mechanism of dynamic stray current and carbon dioxide (CO2) concentration on metallic pipelines was analyzed to present chemical reaction equation of surface corrosion of the metallic pipelines. Multiple linear regression equation was obtained based on simple linear regression equation, then metal pipe surface equation of dynamic stray current and CO2 corrosion was further derived, and the final model for corrosion prediction of metal pipe surface was obtained. Based on the specific examples, the multiple linear regression model for surface corrosion rate prediction of 45 metallic pipelines was simulated by mathematical software MATLAB, and the corrosion rate was compared and analyzed by comparing with the experimental measurement. The corrosion rate of metallic pipeline surface increased with the increase of the dynamic stray current or CO2 concentration. Within 100 h, maximum corrosion rate of metallic pipeline surface with simulated current and CO2 concentration was 3.72×10–4 mm/h and 4.80×10–4 mm/h respectively, and the minimum 3.26×10–4 mm/h and 4.24×10–4 mm/h respectively. Maximum corrosion rate of metallic pipeline surface in current and CO2 concentration experimental measurement was 3.76×10–4 mm/h and 4.86×10–4 mm/h respectively, and the minimum 3.12×10–4 mm/h and 4.08×10–4 mm/h respectively. Meanwhile, the relative error between the theoretical value and the experimental value was less than 5%. The multiple linear regression model can be used to predict corrosion rate of the surface of the metallic gas pipelines approximately, providing reference data for the service life of pipelines and preventing major safety accidents of metallic gas pipelines.