Typical energy audits are sufficiently expensive and time-consuming that many owners and managers of buildings are not willing to invest the time and money required for a full audit. This dissertation provides a methodology to identify buildings with large potential energy savings using limited information, specifically, utility bills, total area and weather data. The methodology is developed based on the hypothesis: if a commercial building is properly designed, constructed, operated, and maintained, the measured energy consumption should approximately match the simulated value for a typical building of the same size with the most efficient HVAC system; otherwise, there may be potential for energy savings. There are four steps in the methodology: 1) testing to determine whether the utility bills include both weather-dependent and weatherindependent loads; 2) separating weather-dependent and weather-independent loads when both are present in the same data; 3) determining the main type of HVAC system; 4) estimating potential energy savings and recommending an energy audit if appropriate. The Flatness Index is selected to test whether the utility bills include both weatherdependent and weather-independent loads. An approach to separate the utility bills based on thermal balance is developed to separate utility bills into weather-dependent and weather-independent loads for facilities in hot and humid climates. The average relative error in estimated cooling consumption is only 1.1% for 40 buildings in Texas, whereas it is -54.8% using the traditional 3P method. An application of fuzzy logic is used to identify the main type of HVAC system in buildings from their 12-month weatherdependent energy consumption. When 40 buildings were tested, 18 systems were identified correctly, seven were incorrect and the HVAC system type cannot be identified in 15 cases. The estimated potential savings by the screening methodology in eight large commercial buildings were compared with audit estimated savings for the same buildings. The audit estimated savings are between 25% - 150% of the potential energy savings estimated by the screening procedure in seven cases. The other two cases are less accurate, indicating that further refinement of the method would be valuable. The data required are easily obtained; the procedure can be carried out automatically, so no experience is required. If the actual type of HVAC system, measured weather-dependent, and weather-independent energy consumption are known, the methodology should work better.