The aims and the usefulness of modelling the thermoregulatory system are outlined by demonstrating applications and results of simulation on different levels of complexity. It is shown that both very simple one-loop models and complex models based on spatially distributed parameters have contributed to a better understanding of the system, but that current issues primarily require the latter type. However, mathematical modelling must be performed in conjunction with experimental studies and must be adapted to the amount of basic physiological data. Future fields of modelling are the adaptive mechanisms and the interactions of systems.