Improving the energy efficiency of buildings in California and beyond has great potential for reducing carbon emissions at low cost. Building energy simulation (ES) tools such as EQuest and EnergyPlus that are used in the design of new and retrofitted energy efficient buildings often fail to accurately predict the energy use in buildings (Turner and Frankel 2008). The more confidence there is that energy simulation tools can accurately model energy use, the more these tools will be utilized in energy efficient design and the more effective performance-based building codes on energy efficiency will be. Building ES tools do not explicitly model the air circulation through the space. As a result, the energy use can be quite challenging to model in buildings with thermal stratification or high airflow velocities, which are particularly common with energy efficient strategies such as displacement ventilation, passive ventilation or mixed-mode ventilation. Moreover, comfort in a building is assessed and controlled at room level, and existing ES methods are particularly poor in characterizing these detailed conditions, leading to inaccurate predictions of the energy use. While sophisticated computational fluid dynamics (CFD) tools exist to model airflow and heat transfer in complex spaces, the application of these tools to building-energy problems is limited, given the cost of implementation and simulation. We propose to develop a building energy model using the current best practices in CFD, coupled with heat transfer information from existing an existing energy simulation tool. The intention is not to develop a replacement for existing ES tools, but instead to use an integrated CFD-ES model to identify conditions where discrepancies occur between standard ES predictions and actual energy use in buildings. The extensive building monitoring data from the Yang and Yamazaki Environment and Energy (Y2E2) building and the Santa Clara County Jail building offer a unique opportunity to validate this coupled model. Identifying conditions under which energy models are underperforming would help to focus the future development of building energy modeling.