Abstract As two of the most important products of the combustion process, carbon dioxide (CO2) and carbon monoxide (CO) are commonly used as tracers for combustion source assignment. Their relationship will help to better understand the regional carbon cycle and assess climate forcing effects. In this study, mixing ratios of CO2 and CO were continuously measured using a Picarro gas concentration analyzer at the Atmospheric Boundary Layer Eco-Environmental Shanghuang Observatory, Chinese Academy of Sciences (ABLECAS) throughout 2022–2023. The variability of the mixing ratio of CO to CO2 (ΔCO/ΔCO2) in a 1 h time interval was calculated based on linear slope analysis after background values were determined and subtracted. The results showed that the mixing ratio of CO had a clear seasonal variability with a moderate increase in the spring (249.1 ± 59.6 part per billion (ppb)) and winter (257.8 ± 90.3 ppb), mostly due to more frequent transport from north of the Yangtze River. ΔCO/ΔCO2 at the ABLECAS varied with air mass origin, with a linear slope 0%–1% on a 1 h basis. Relatively high ΔCO/ΔCO2 values for an air mass from the north in the winter indicate that the emission sources had lower combustion efficiency. In summer, the ΔCO/ΔCO2 ratio mostly reflected the background conditions for air masses from marine areas. The potential source regions and contribution assignments were evaluatedat the ABLECAS according to source–receptor relationship analysis using the FLEXPART model with CO as a pollutant tracer from 2015 to 2023. We found that the footprint of an air mass had a clear transition period between 2018 and 2019, and a synoptic anomaly, related to Arctic Oscillation strength and west Pacific subtropical high position, plays a key role in influencing the pollutant transport patterns. This study provides a scientific basis for the formulation of air quality regulation policy, and helps to implement the national carbon neutralization strategy.