This study analyzed systematic biases in surface air temperature (SAT) within Far East Asia during the boreal winter using the SNURCM and WRF regional climate models (RCMs) from the Coordinated Regional Downscaling Experiment (CORDEX)-East Asia phase II. The SAT biases were examined in relation to the East Asian winter monsoon (EAWM). The models consistently simulated lower winter temperatures over East Asia, particularly in the Manchuria (MC) region, compared to the observation, showing a positive correlation with the EAWM. This study assessed the models' ability to capture EAWM variability and revealed relationships between SAT biases and discrepancies in low-level and near-surface EAWM conditions. The findings emphasized the value of analyzing extreme monsoon years, with the RCMs exhibiting larger cold SAT biases during strong EAWM years. Systematic biases in sea-level pressure contrast and lower-level winds over the MC region were evident during years with a robust monsoon. The overestimation of low-level winds during strong EAWM years contributed to increased cold advection, affecting the MC region. These systematic errors are influenced by the internal factors of the model, such as the physics parameterization schemes, rather than large-scale circulation forced by the reanalysis data (perfect boundary condition). These results provide insights for model improvements, understanding EAWM dynamics, and call for investigation of processes in the planetary boundary layer and coupled air-sea interaction.