In order to evaluate health risks under severe thermal environments, we are developing a new human thermophysiological model, which can predict not only body temperature but also blood pressure (BP) and blood flow rate (BFR). In the present study, we developed a sub-model (thermal network model) to simulate body temperature, and carried out subjective experiments under different ambient temperatures to collect reference data. The experiments indicated that human thermoregulatory ability in cold environments is weaker than that in hot environments. The experiments also showed that BFRs of limbs compose only small portions of total BFR but increase greatly with ambient temperature. On the other hand, our thermal network model could reproduce the trend of tympanic temperature and mean skin temperature. In addition, some numerical simulations showed that the BFRs of limbs are important for thermoregulation even if their portions are small, and besides, heat transfer along vessels is also important.