Global land surface models (LSMs) simulate the interaction between the terrestrial biosphere and the climate system. These models are an uncertain component of Earth System Models that underpin the projections of future climate. LSMs have been extensively validated at short time scales (i.e., days to years) but remain poorly constrained at the time scales of global change from decades to centuries. The search for long-term model benchmarks has brought tree-ring data to the attention of the modelling community. Tree rings can constrain simulations of biomass growth and long-term changes in tree-level gas exchange, a process that underlies photosynthesis and transpiration. In this talk I will present ideas and recent advances for benchmarking the French ORCHIDEE global land surface model (https://orchidee.ipsl.fr/) with tree-ring widths and their stable isotopic composition.