Abstract. To understand the carbon cycle at policy-relevant spatial scales, a high density of high-quality CO2 measurement sites is needed. In 2012, the Korea Meteorological Administration (KMA) installed CO2 monitoring systems at Anmyeondo (AMY) in the west, Jejudo Gosan Suwolbong (JGS) in the southwest, and Ulleungdo (ULD) in the east of South Korea. Three stations were instrumented with identical greenhouse gas measurement systems based on cavity ring-down spectroscopy (CRDS) and a new drying system developed by KMA and the Korea Research Institute of Standards and Science (KRISS). This drying system is suitable in humid areas; water vapor measured using CRDS in ambient air was 0.001 % to 0.004 % across the stations. Measurement uncertainties expressed by the quadrature sum of the uncertainties from the drying system, scale propagations, repeatability, and reproducibility were ∼0.11 ppm from all KMA stations in the 68 % confidence interval. Average monthly CO2 enhancements above the local background at each station were 4.3±3.3 ppm at AMY, 1.7±1.3 ppm at JGS, and 1±1.9 ppm (1σ) at ULD, respectively, during 2012 to 2016. At AMY station, located between China and South Korea, CO2 annual means and seasonal variations are also greater than the other KMA stations, indicating that it is affected not only by local vegetation, but also added anthropogenic sources. Selected baseline CO2 at AMY and at JGS in the west of South Korea is more sensitive to East Asia (e.g., China) according to wind direction and speed. Through the comparison of long-term trends and growth rates at AMY with other East Asian stations over 15 years, it was suggested that they could be affected not only by local vegetation but also by measurement quality.