This session focuses on efforts to improve our understanding of how global vegetation will respond to changes in atmospheric CO2. Specifically, we are looking at efforts to benchmark process-based models against long-term experimental data from the Free Air CO2 Enrichment (FACE) experiments. The talks in this session highlight the improvements made in these models as a result of integration with field experiments as well as the novel insights the models can provide about the results from these experiments. The goals of this session are directly in line with this year's theme for ESA which focuses on global warming as a challenge for the future. The models presented in this session are actively being used to forecast how vegetation will respond to climate change at regional to global scales, including some that can be linked to GCMs and are part of the IPCC reports. The IPCC reports have demonstrated substantial uncertainty in the feedbacks between ecosystems and the atmosphere that determine the long-term trajectory of atmospheric CO2 and global temperature. These models have often made simplifying assumptions about CO2 fertilization effects that are consistent with elevated CO2 research but until now the models had not been challenged to capture the full-range of data from FACE experiments. These tests are particularly important given the growing recognition of the importance of feedbacks such as progressive nitrogen limitation. The core of this session will present results from a NCEAS-sponsored model-data intercomparison project to test a dozen models at two forest FACE sites. These talks will focus on the connections to different biogeochemical cycles (Oren, Parton, Thorton) and the different temporal scales (Kicklighter). We will also highlight the techniques used to bring models and data together (Luo). Additional talks in the session will expand the domain considered to other biomes (Ogle, Twain) and larger spatial scales(Zaehle, Jain, Prentice). The scope of our session is not limited to modelers but aims to highlight what the modeling and field research communities can learn from each other and to clarify what types of processes need to be better understood, and what measurements are currently lacking but are needed in order to improve our ability to predict changes in the global carbon cycle.