Calculation and measurement-based energy performance evaluations of the same building often provide different results. This difference is referred as "the performance gap". However, a large performance gap may not necessarily mean that there are flaws in the building or deviations from the intended design. The causes for the performance gap can be analysed by calibrating the simulation model to measured data. In this paper, an approach is introduced for verifying compliance with energy performance criteria of residential buildings. The approach is based on a performance gap analysis that takes the uncertainties in the energy evaluation methods into consideration. The scope is to verify building energy performance through simulation and analysis of measured data, identifying any performance gap due to deviations from the intended design or flaws in the finished building based on performance gap analysis. In the approach, a simulation model is calibrated to match the heat loss coefficient of the building envelope [kWh/K] instead of the measured energy. The introduced approach is illustrated using a single-family residential building. The heat loss coefficient was found useful towards identifying any deviations from the intended design or flaws in the finished building. The case study indicated that the method uncertainty was important to consider in the performance gap analysis and that the proposed approach is applicable even when the performance gap appears to be non-existing.