As people spend most of their time indoors, steady state models may be appropriate for thermal comfort assessments there, while, for the relatively short times spent outdoors (mostly less than 1 h), those models tend to overestimate discomfort. This discrepancy is larger for cool outdoor conditions than for warm ones. For example, for a person leaving a room in thermal comfort (T-sk = 33.5 degreesC, T-core = 37 degreesC) into cold winter outdoor conditions (T-a = T-mrt = 0degreesC, VP = 5 hPa, v = 1 m/s) obtaining steady state will take many hours, while leaving into hot conditions (T-a = 30 degreesC, T-mrt = 60 degreesC, VP = 15 hPa, v = 0.5 m/s) it will be reached within less than 30 min. Consequently, especially for outdoor thermal comfort assessments in cold conditions, non-steady state models should be applied. This, among other new aspects, will be considered in a new internationally standardised Universal Thermal Climate Index (UTCI), which is currently being developed by a working group of the International Society of Biometeorology. Besides such physical/physiological approaches, psychological factors also have to be considered, i.e. diverging thermal expectations indoors and outdoors. Consequently, different approaches are necessary for assessing indoor or outdoor thermal comfort.