Previous studies have shown that the choice of the behavioural model for the control of lighting and shading devices has important implications in the energy performance assessment of office buildings and directly influences the comparison of design alternatives for control systems as well as for the façades (shading or glazing). Since many behavioural models exist for the same purposes, and in most cases there is no consensus on which ones are more representative of the reality, it is important to perform more real-world observations and draw lessons from them. This study appears in this context and explores the results of a field campaign that continuously monitored occupants’ behaviour regarding manual control of electric lighting and shading devices of eight real-world single offices under normal operation. Detailed measurements and observations were performed over two months in each office with the aim of assessing the level of agreement between existing behavioural models and actual occupants’ behaviour. Findings from this study enabled refining the hypothesis regarding realistic modelling of occupants’ behaviour when considering currently available whole building simulation (BS) software; moreover, needed improvements to increase the reliability of the results in whole BS were identified.