Measurements of 18O in atmospheric CO2 can be used to trace gross photosynthetic and respiratory CO2 fluxes between the atmosphere and the terrestrial biosphere. However, this requires knowledge of the 18O signatures attributable to the fluxes from soil and leaves. Newly developed methods were employed to measure the 18O of soil-respired CO2 and depth profiles of near-surface soil CO2, in order to evaluate the factors influencing isotopic soil-atmosphere CO2 exchange. The 18O of soil-respired CO2 varied predominantly as a function of the 18O of soil water which, in turn, changed with soil drying and with seasonal variations in source water. The 18O of soil-respired CO2 corresponds to full isotopic equilibrium with soil water at a depth ranging between 5 and 15 cm. The 18O of respired CO2, in reality, results from a weighted average of partial equilibria over a range of depths. Soil water isotopic enrichment of up to 10? in the top 5 cm did not appear to strongly influence the isotopic composition of the respired CO2. We demonstrate that during measurements invasion of atmospheric CO2 (the diffusion of ambient CO2 into the soil, followed by partial equilibration and retrodiffusion) must be considered to accurately calculate the 18O of the soil-respired CO2. The impact of invasion in natural settings is also considered. We also have determined the effective kinetic fractionation of CO2 diffusion out of the soil to be 7.2¿0.3?. High-resolution (1 cm) depth profiles of 18O of near-surface (top 10 cm) soil CO2 were carried out by gas chromatography-isotope ratio mass spectrometry (GC-IRMS). This novel technique allowed us to observe the competitive diffusion-equilibration process near the soil surface and to test simulations by a diffusion and equilibration model of the soil CO2 18O content. ¿ 1999 American Geophysical Union