Rising levels of atmospheric CO2 concentration (Ca) and simultaneous climate change profoundly affect plant physiological performance while challenging our ability to estimate vegetation–atmosphere fluxes. To predict rates of water and carbon exchange between vegetation and the atmosphere, we require a formulation for stomatal conductance (gs) that captures the multidimensional response of stomata to changing environmental conditions. The unified stomatal optimization (USO) theory provides a formulation for gs with the ability to predict the response of gs to novel environmental conditions such as elevated Ca (eCa), warmer temperatures and/or changing water availability. We tested for the effect of eCa and seasonally varying climate on stomatal behaviour, as defined by the USO theory, during the first year of free-air CO2 enrichment in a native eucalypt woodland (the EucFACE experiment). We hypothesized that under eCa, gs would decrease and photosynthesis (Anet) would increase, but fundamental stomatal behaviour described in the USO model would remain unchanged. We also predicted that the USO slope parameter g1 would increase with temperature and water availability. Over 20 months, we performed quarterly gas exchange campaigns encompassing a wide range of temperatures and water availabilities. We measured gs, Anet and leaf water potential (Ψ) at mid-morning, midday and pre-dawn (Ψ only) under ambient and eCa and prevailing climatic conditions, at the tree tops (20 m height). We found that eCa induced a 20% reduction in stomatal conductance under non-limiting water availability, enhanced mid-morning Anet by 24% in three out of five measurement campaigns and had no significant effect on Ψ. The parameter g1 was conserved under eCa, weakly increased with temperature and did not respond to increasing water availability. Our results suggest that under eCa and variable rainfall, mature eucalypt trees exhibit a conservative water-use strategy, but this strategy may be modified by growth temperature. We show that the USO theory successfully predicts coupling of carbon uptake and water loss in future atmospheric conditions in a native woodland and thus could be incorporated into ecosystem-scale and global vegetation models.Read less <