Abstract. In China's coal producing areas, carbon emission has a significant impact on the environment, so more attention has focused on the development of heavy vehicles and expressway designation. However, no national criterion based on carbon emission for overload expressway designation exists. In this paper, we propose building a connection between longitudinal segment design parameters and carbon emission, using diesel fuel consumption data and a nonlinear iterative regression method. A carbon emission relational model was built, showing the connection between carbon emission and gradient, slope length, and initial speed. The carbon emission database of expressway longitudinal slope was established by existing model. Based on the actual design data, the sensitivity analysis of gradient and slope length in the established model is carried out. As a result, the influence of gradient on the cumulative carbon emissions is obvious and higher sensitivity, reaching 4.550. The carbon emissions will increase significantly after the longitudinal slope is greater than 3%, highway longitudinal slope should be less than 3%, and the road longitudinal slope length should be limited.