AbstractVegetation exerts large control on global biogeochemical cycles through the processes of photosynthesis and transpiration that exchange CO2 and water between the land and the atmosphere. Increasing atmospheric CO2 concentrations exert direct effects on vegetation through enhanced photosynthesis and reduced stomatal conductance, and indirect effects through changes in climatic variables that drive these processes. How these direct and indirect CO2 impacts interact with each other to affect plant productivity and water use has not been explicitly analysed and remains unclear, yet is important to fully understand the response of the global carbon cycle to future climate change. Here, we use a set of factorial modelling experiments to quantify the direct and indirect impacts of atmospheric CO2 and their interaction on yield and water use in bioenergy short rotation coppice poplar, in addition to quantifying the impact of other environmental drivers such as soil type. We use the JULES land‐surface model forced with a ten‐member ensemble of projected climate change for 2100 with atmospheric CO2 concentrations representative of the A1B emissions scenario. We show that the simulated response of plant productivity to future climate change was nonadditive in JULES, however this nonadditivity was not apparent for plant transpiration. The responses of both growth and transpiration under all experimental scenarios were highly variable between sites, highlighting the complexity of interactions between direct physiological CO2 effects and indirect climate effects. As a result, no general pattern explaining the response of bioenergy poplar water use and yield to future climate change could be discerned across sites. This study suggests attempts to infer future climate change impacts on the land biosphere from studies that force with either the direct or indirect CO2 effects in isolation from each other may lead to incorrect conclusions in terms of both the direction and magnitude of plant response to future climate change.