Energy consumption of residential buildings and offices adds up to about 30% of total carbon dioxide emissions; and occupant behaviour contributes to 80% of the variation in energy consumption. Indoor climate regulations are based on an empirical thermal comfort model that was developed in the 1960s (ref. ). Standard values for one of its primary variables--metabolic rate--are based on an average male, and may overestimate female metabolic rate by up to 35% (ref. ). This may cause buildings to be intrinsically non-energy-efficient in providing comfort to females. Therefore, we make a case to use actual metabolic rates. Moreover, with a biophysical analysis we illustrate the effect of miscalculating metabolic rate on female thermal demand. The approach is fundamentally different from current empirical thermal comfort models and builds up predictions from the physical and physiological constraints, rather than statistical association to thermal comfort. It provides a substantiation of the thermal comfort standard on the population level and adds flexibility to predict thermal demand of subpopulations and individuals. Ultimately, an accurate representation of thermal demand of all occupants leads to actual energy consumption predictions and real energy savings of buildings that are designed and operated by the buildings services community.