delta(18)O in CO2 attracts attention for its ability to decipher the gross fluxes of terrestrial CO2 exchange. This capacity can only be exploited if one knows the impact of the biology and the water isotope cycle on the isotopic composition of atmospheric CO2. To study the processes, we built an integrated global model of delta(18)O in atmospheric CO2 that calculates O-18 in the water cycle pools, the CO2 and the inherent delta(18)O-CO2 fluxes and transports these in the atmosphere. Within the framework of the European Project EUROSIBERIAN CARBONFLUX, we investigated the O-18 processes on the continent of Eurasia north of 40 degreesN. We show that there is a large impoverishment in the water isotopic composition of rain in the continental interior but this is significantly reduced during the growing season. We validated the model directly by comparing it to Net Ecosystem Exchange (NEE) measurements from eddy-flux towers and indirectly by comparing it to atmospheric measurements Of CO2 and delta(18)O-CO2 at 3000 m a.s.l. The model reproduces well the seasonal Cycle Of CO2 and the phase for delta(18)O-CO2 but it underestimates the amplitude of that species. Our model further predicts a large diurnal rectifier effect that is different in CO2 than in delta(18)O-CO2 in eastern Siberia due to very low or even negative leaf discrimination. We analyse the longitudinal gradient in leaf discrimination which reflects primarily the gradient in leaf water isotopic composition. The leaf water delta(18)O-H2O gradient is itself determined 25% by the meridional gradient in source water and 75% by the change of relative humidity with longitude.