Below-canopy eddy covariance measurements of CO2 flux (Fcb) and soil surface CO2flux measurements (Fs) were made seasonally in a ponderosa pine forest in central Oregon in 1996 and 1997. The forest ecosystem has a very open canopy, and it is subject to drought and high vapor pressure deficits in summer. Below-canopy flux measurements in March, May, and August 1997 showed increasing effluxes from the forest floor as soils warmed. In July 1996, daytime Fcb measurements appeared to have been influenced by photosynthetic uptake of CO2 by ground vegetation. We did not see a similar diurnal trend in Fcb data in August 1997, probably because photosynthesis may have decreased with senescence of ∼1/3 of the pine canopy and the herbaceous species. On 4 days in August 1997, the mean nocturnal Fs (2.6 ± 0.08 μmol m−2 s−1) was lower than nocturnal Fcb (3.5 ± 0.28 μmol m−2 s−1) by 26%, and daytime Fs was lower than nocturnal Fcb by 18%, possibly because Fcb includes respiration by understory and the lower portions of trees. The mean nocturnal NEE calculated from above-canopy flux and storage in the canopy airspace (Fca + Fstor) at this time was 2.8 ± 0.40 μmol m−2 s−1, 23% lower than ecosystem respiration calculated from chamber measurements on soils, wood, and foliage. The largest difference was observed on a more a turbulent night (u* = 0.30 m s−1) when Fca + Fstor was even significantly less than Fcb and Fs. Our hypothesis is that under calm conditions (e.g. u* < 0.15 m s−1 as observed on three of the nights), Fca is negligible and has no impact on the CO2 budget. Under weak wind conditions (e.g. u* = 0.30 m s−1), Fca begins to become significant and fluxes missed by the above-canopy eddy correlation system degrade the CO2 budget. Under windy conditions, the above-canopy eddy correlation measurement is a good approximation and the CO2 budget improves again. Below-canopy flux measurements provided useful temporal information for understanding seasonal differences in diel patterns, while the chambers allowed us to characterize spatial variation in CO2 fluxes. It is important to measure below-canopy fluxes along with above-canopy fluxes throughout the year to understand CO2 exchange components and annual contributions to the carbon budget of open canopy forest systems.