Background/Question/Methods Forests are one of the largest conduits in the transfer of carbon from Earth’s atmosphere to the terrestrial biosphere, accounting for nearly half of terrestrial net primary productivity (NPP). However, NPP represents only a fraction of the total amount of carbon fixed during photosynthesis (GPP), as a substantial amount of CO2 is released back to the atmosphere through respiration. Future changes in tree physiology driven by elevated CO2 (eCO2) or climate change may not affect carbon uptake and respiration equally and may impact the ratio of NPP to GPP, defined as tree carbon use efficiency (CUE), with important implications for the global carbon cycle. Here, we update an optimality-based tree model to include temperature sensitivity to photosynthesis and respiration to examine the consequences of future environmental change on tree CUE globally. We first validate model predictions using a suite of CUE observations from trees around the world and six different representations of respiration sensitivity to temperature commonly utilized in Earth-system models. We then examine the effects of eCO2 and climate on simulated CUE using climate projections from the Coupled Model Intercomparison Project phase 5 (CMIP5) from seven different coupled climate-Earth system models. Results/Conclusions We show the updated model accurately partitions carbon assimilated to NPP and respiration, with a global median CUE of 0.43 across all grid-cells containing forested area, similar to the global median of examined tree CUE observations from published literature (0.46, N = 228), and well within the total range of tree CUE observations (0.22 – 0.79). Moreover, we show simulated CUE declines with increasing temperature and falls within the interquartile range of all CUE observations examined across a range of temperature (12 – 28°C), as respiration increases proportionally more than photosynthesis. Averaged across all respiration functions, simulated tree CUE increases by 0.025 for every 100 ppm increase in atmospheric CO2, but declines by 0.014 for every 1°C increase in temperature, resulting in a net increase of CUE by 0.027 globally between the historical and future periods. Last, we found that model-predicted carbon balance based on local climate conditions was prognostic of forest regions as diagnosed by lidar-derived forest canopy height; as canopy height decreased below 7 meters the model predicted unsustainable conditions for tree growth in the majority of locations. Thus, we use simulated carbon balance to understand potential shifts in forest distribution under end-of-century climate conditions and eCO2.