Energy modeling has become widely utilized in the design process of buildings due to rating programs such as Leadership in Energy and Environmental Design (LEED), which require the use of modeling programs to evaluate potential energy efficiency. The accuracy of these models, however, has been under much scrutiny. Some LEED buildings have even been documented as consuming more energy than conventional buildings. If building energy cannot be accurately predicted, then the efficacy of energy modeling will be in question and broader implications, including the value of LEED, may become pervasive. This research analyzes the results of design phase energy models with respect to actual energy usage. Energy data was collected for 16 months in the case study building, a Solar House. The measured data was compared with results from three different energy models - Green Building Studio, Energy-10, and EnergyPlus. The average error rate between the measured and predicted energy usage was 59%. Discrepancies between the models and actual data included occupant activities and system efficiencies.