CO2 emission is increasingly focused by public. Beijing and Tianjin are conceived to be a new economic point of growth in China. However, both of them are suffering serious environmental stress. In order to seek for the effect of socioeconomic factors on the CO2 emission of this region, a novel methodology âsymbolic regressionâ is adopted to investigate the relationship between CO2 emission and influential factors of Beijing and Tianjin. Based on this method, CO2 emission models of Beijing and Tianjin are built respectively. The models results manifested that Beijing and Tianjin own different CO2 emission indicators. The RMSE of models in Beijing and Tianjin are 255.39 and 603.99, respectively. Further analysis on indicators and forecast trend shows that CO2 emission of Beijing expresses an inverted-U shaped curve, whilst Tianjin owns a monotonically increasing trend. From analytical results, it could be argued that the diversity rooted in different development orientation and the mixture of different natural and industrial environment. This research further expands the investigation on CO2 emission of Beijing and Tianjin region, and can be used for reference in the study of carbon emissions in similar regions. Based on the investigation, several policy suggestions are presented. Keywords: Beijing and Tianjin, CO2 emission, Driven factors, Genetic programming, Pareto front, Symbolic regression