Evolution of energy standards led to high-performance buildings requiring very low energy for their operation. Occupancy is the variable with the greatest impact on nearly Zero Energy Building (nZEB) energy performance because both constructive and technical characteristics have been improved over time. Occupants influence energy use in buildings as they contribute to internal gains, interact with systems and modify indoor conditions with their behaviour. Assumptions about occupancy schedules are usually adopted in energy models for compliance calculation and when experimental data are not available. These theoretical profiles might be far from real conditions and frequently generate a mismatch between expected and actual performance. The present work analyses six months of monitored data from an nZEB in Denmark. A simulation model is used to analyse the effect of three different occupancy profiles on the final energy use: the ‘Compliance profile’, defined on the basis of regulations, the ‘Standard profile’, built on average data obtained from surveys, and the ‘Actual profile’, customized on measured data from the actual building case. Significant differences are detected in the three different occupancy profiles as well as in the results achieved by applying the three occupancy models in performance prediction.