Today's inefficient operation of HVAC (Heating Ventilation Air Conditioning) systems causes insufficient thermal comfort and increases operating costs of such systems. Continuous evaluation of HVAC systems performance would provide building operators with the feedback needed to run the HVAC systems much more efficiently and effectively than today. Our vision is to automate the generation of this feedback as much as possible based on comparing predicted and measured HVAC system performance data. This ongoing research (sponsored by CIFE, LBNL and PEEC) we have already developed an overall methodology to compare predicted and measured data. In applying the methodology on two test cases (SF Federal Building and Global Ecology Building) we realized that we need to understand the conditions that make such a comparison possible at a much more fundamental level than exists today. Thus we are working on two more case studies (Y2E2 and a Building in Santa Clara County) to investigate the underlying assumptions and possible sources of modeling and measurement errors. The research also .../includes the development of guidelines for the modeling and simulation and the measurement of the energy performance of HVAC systems for this comparison.