Windows are one of the most complex determinants of the office indoor environment and have a significant impact on the overall energy demand for electric lighting, heating and cooling, which typically account for more than 50% and often more than 70% of the final energy demand of office buildings. Building occupants tend to adjust the lighting and shading devices dynamically, as a function of the indoor and outdoor environmental conditions and of the way how these conditions impact their visual comfort perception. Several control criteria to characterize this interaction have being proposed and some of those criteria are already integrated in building simulation tools. This research aimed to assess the impact of considering a given control model instead of another on predicted energy demand, and to evaluate how consistent the different behavioral models suggested in the literature and currently used in dynamic simulation are when compared with the patterns of control observed in real office environments. A comprehensive simulation case study carried out to understand the impact of the use of different control patterns on the selection of transparent façade design options revealed that these influence significantly the predicted energy performance and also the selection of design alternatives. In order to obtain some field indications of how occupants actually control their lighting and shading devices and, moreover, how the observed behaviors match the patterns published in the literature, eight single-occupant offices were monitored during two months each. Monitoring results showed that the actions of controlling the lighting state based solely on the availability of daylight are less than those associated with established schedules or routines. However, daylight availability does have an impact on the decision to turn lighting on when arriving to the office, at least when returning from long absences. It was also observed that the action of closing the shadings during occupation time are primarily connected to high values of transmitted solar radiation or direct solar radiation hitting the workplace. The results of this research indicated some implications regarding the use of buildings energy simulation softwares and the control/behavioral models that they currently include. In particular, it seems that the current algorithms to control the electric lighting should be used only in connection with daylight controls. The main needs for further developments in the control of shadings regards the inclusion of direct solar radiation as a decision variable, and a modulation according to DGI and/or DGP thresholds when opening the shadings.