Abstract: Air temperature is a key factor affecting the natural gas load in the heating season. Based on the historical data of daily gas consumption and air temperature in the heating season in Hebei Province, the correlation degree is compared by calculating the correlation coefficient between the gas consumption and the daily average air temperature, maximum air temperature and minimum air temperature. The regression analysis and binary regression analysis are performed respectively.Regression equation is obtained to predict the gas consumption in the next heating season without major policy changes. By comparison, the fitting and forecasting results of the binary regression are better. Considering the influence of the accumulation effect of air temperature on the gas consumption in the heating season, the temperature is corrected by using different air temperature accumulation effect coefficients selected according to the different daily air temperature. The calculation results show that the accuracy is improved after temperature is corrected. In general, this study provides a reference for more accurate prediction of natural gas load by analyzing the relationship between natural gas load and air temperature during the heating season.