In order to maximise occupants’ well-being and work efficiency, keep satisfying environmental conditions, and minimise costs and impact related to the operative and energy use, a “human-centric” approach is more and more pursued in buildings’ design and performance assessment. In this context, the use of Virtual Reality (VR) is emerging due to its advantages (low cost, repeatability, and speed of execution) compared to physical study settings. However, in Immersive Virtual Environments (IVE), it is important to ensure that data represented and collected faithfully replicate the physical environments. In order to provide a further contribution in terms of IVE validation process in the building field, this research presents results from an experimental study, where subjects’ performance tests and comfort assessments were compared in real and virtual office settings under three different walls colour layouts and two air temperature levels. “Internal”, “ecological” and “construct” validity of the IVE have been demonstrated. Findings revealed no statistically significant differences in productivity and sensation votes and in the impact of colour and temperature variables. Results then highlight a strong agreement of the two tested environments, revealing that VR is a potentially reliable tool to measure its real counterparts in terms of occupants’ productivity, perception, and behaviour under different test conditions.