Abstract: In alignment with the national carbon peaking and carbon neutrality goals, this study introduces a Graded Variable Speed Limit strategy to enhance the level of low-carbon services in highway transportation. The strategy focuses on controlling speed limits in bottleneck areas of highways under mixed traffic flow scenarios, which include connected and automated vehicles (CAV) and human driven vehicles (HDV). Capitalizing on the controllability advantages of CAVs, this strategy refines the traditional variable speed limits (VSL) control model by incorporating hierarchical levels and safety reduction intervals. This addresses issues such as the inflexibility and slow response of existing VSL strategies. The control model aims to minimize the total travel time (TTT) of vehicles within the controlled section and environmental indicators (NO_x, CO) while ensuring travel efficiency. Considering the distinct traffic flow characteristics and interaction features between CAVs and HDVs in mixed traffic states, a simulation environment is established to emulate the operational conditions of mixed traffic flow in real-world scenarios. The effectiveness of the control model is validated through comparative experiments. The results indicate that with a 40% CAV penetration rate (CAVs comprise 40% of the total vehicular composition), a mere 0.92% reduction in traffic flow leads to significant improvements in various low-carbon environmental indicators. Specifically, CO emissions, NO_x emissions, PM_x emissions, HC emissions, CO₂ emissions, and average fuel consumption decrease by 33.33%, 15.36%, 19.97%, 27.32%, 13.01%, and 13.01%, respectively. The research findings contribute to further advancing the “low-carbon” transformation of highways under mixed traffic flow environments, serving the carbon peaking and carbon neutrality goals. To a certain extent, a balance is achieved between travel efficiency and carbon emission control, realizing an equilibrium between control effectiveness and construction costs.