Abstract: To study the internal corrosion of ground pipelines in Yuanba high-sour gasfield, a corrosion prediction model based on BP neural network was proposed with consideration to multiple factors. In the model, the input was 8 types of working condition data, i.e., the temperature, CO₂ partial pressure, H₂S partial pressure, pH value, Cl- concentration, total salinity, liquid-to-gas ratio and the residue of corrosion inhibitor, while the output was the corrosion rate. Specifically, the BP neural network was trained with a large amount of historical sample data from field measurement to realize the prediction of the corrosion rate of ground pipelines. Besides, the model was used to evaluate the importance of various factors affecting the corrosion of ground pipelines in Yuanba high-sour gasfield. As shown in the results, the predicted model is highly accurate and reliable, as the average absolute error between the predicted value of the model and the measured value is within 10% for random working condition parameters. In addition, H₂S partial pressure is the main control factor of the corrosion rate of ground pipelines in Yuanba high-sour gasfield, followed by CO₂ partial pressure and corrosion inhibitor. Generally, the research results could provide technical reference for the internal corrosion assessment of ground pipelines in similar gas fields.