This paper introduces a longitudinal study monitoring occupants’ window opening behaviour in a mixed-mode office building in Beijing, China, when natural ventilation is specifically used for controlling the building’s indoor thermal environment. Based on the field measured data, the influence of factors, including outdoor air temperature, outdoor PM2.5, indoor air temperature, time of day, occupancy and previous window state, on the observed state of windows is analysed. All of them are influential on occupants’ window opening behaviour in the case study building, and so they can be used to model occupants’ window opening behaviour in buildings in China to achieve a better consideration of occupant behaviour in dynamic building performance simulation.