We examined potential interferences from water vapor and atmospheric backgroundgases (N2, O2, and Ar), and biases by isotopologues of target species, on accuratemeasurement of atmospheric CO2 and CH4 by means of wavelength-scanned cavity5 ring-down spectroscopy (WS-CRDS). Variations in the composition of the backgroundgas substantially impacted the CO2 and CH4 measurements: the measured amounts ofCO2 and CH4 decreased with increasing N2 mole fraction, but increased with increasing O2 and Ar, suggesting that the pressure-broadening effects (PBEs) increased asAr < O2 < N2. Using these experimental results, we inferred PBEs for the measurement10 of synthetic standard gases. The PBEs were negligible (up to 0.05 ppm for CO2 and0.01 ppb for CH4) for gas standards balanced with purified air, although the PBEs weresubstantial (up to 0.87 ppm for CO2 and 1.4 ppb for CH4) for standards balanced withsynthetic air. For isotopic biases on CO2 measurements, we compared experimentalresults and theoretical calculations, which showed excellent agreement within their un-15 certainty. We derived empirical correction functions for water vapor for three WS-CRDSinstruments (Picarro EnviroSense 3000i, G-1301, and G-2301). Although the transferability of the functions was not clear, no significant difference was found in the watervapor correction values among these instruments within the typical analytical precisionat sufficiently low water concentrations (< 0.3 %V for CO2 and < 0.4 %V for CH4). Foraccurate measurements of CO2 and CH420 in ambient air, we concluded that WS-CRDSmeasurements should be performed under complete dehumidification of air samples,or moderate dehumidification followed by application of a water vapor correction function, along with calibration by natural air-based standard gases or purified air-balancedsynthetic standard gases with isotopic correction.