Increasing temperatures and atmospheric CO sub(2) concentrations will affect tree carbon fluxes, generating potential feedbacks between forests and the global climate system. We studied how elevated temperatures and CO sub(2) impacted leaf carbon dynamics in Norway spruce (Picea abies), a dominant northern forest species, to improve predictions of future photosynthetic and respiratory fluxes from high-latitude conifers. Seedlings were grown under ambient (AC, c. 435 mu mol mol super(-1)) or elevated (EC, 750 mu mol mol super(-1)) CO sub(2) concentrations at ambient, +4 degree C, or +8 degree C growing temperatures. Photosynthetic rates (A sub(sat)) were high in +4 degree C/EC seedlings and lowest in +8 degree C spruce, implying that moderate, but not extreme, climate change may stimulate carbon uptake. A sub(sat), dark respiration (R sub(dark)), and light respiration (R sub(light)) rates acclimated to temperature, but not CO sub(2): the thermal optimum of A sub(sat) increased, and R sub(dark) and R sub(light) were suppressed under warming. In all treatments, the Q sub(10) of R sub(light) (the relative increase in respiration for a 10 degree C increase in leaf temperature) was 35% higher than the Q sub(10) of R sub(dark), so the ratio of R sub(light) to R sub(dark) increased with rising leaf temperature. However, across all treatments and a range of 10-40 degree C leaf temperatures, a consistent relationship between R sub(light) and R sub(dark) was found, which could be used to model R sub(light) in future climates. Acclimation reduced daily modeled respiratory losses from warm-grown seedlings by 22-56%. When R sub(light) was modeled as a constant fraction of R sub(dark), modeled daily respiratory losses were 11-65% greater than when using measured values of R sub(light). Our findings highlight the impact of acclimation to future climates on predictions of carbon uptake and losses in northern trees, in particular the need to model daytime respiratory losses from direct measurements of R sub(light) or appropriate relationships with R sub(dark).