Among the various influential factors that impact building performance, occupant behaviour is considered one of the most important factors. Occupants dynamically interact with various indoor environment control systems, such as thermostats, operable windows and blinds, to perform daily routines and/or to maintain, restore, or improve their comfort and needs in an indoor environment. Conventional survey approaches have offered valuable insights into changes in occupants’ attitudes, behaviours, and the characteristics of sample populations, either over time or at a single point in time. However, studies exploring occupants’ preferences for using various building controls in situations involving multiple types of discomfort (e.g., thermal, visual, or acoustic) remain limited. This study incorporates a discrete choice experiment to obtain occupants’ perspective on building usage. In addition to a conventional survey, it also collects occupants’ choice preferences for energy-related comfort adaptive behaviour in office settings. The experiment design includes four types of influential factors: building characteristics, contextual factors, environmental conditions, and personal attributes. Mixed logit regression is used to reveal the utilities and to generate probability model from the data. The results show that occupant behaviours in office environments are strongly influenced by tasks, location, and expected social interactions, while also demonstrating a cooperative intention among co-workers. This model complements existing occupant behaviour models and agent-based modelling approaches by accounting for multiple behaviours (e.g., window opening and blinds use) and the capability to model both individual and collaborative behaviour.