Physical restraint of rodents is needed for nose-only exposure to airborne toxicants and is also used as a means of psychological stress. Hyperthermia is often observed in restrained rats, presumably as a result of impairments in heat dissipation. However, such a hyperthermic response should be dependent on the prevailing ambient conditions. To understand how ambient temperature (Ta) affects the thermoregulatory response to restraint, core temperature (Tc) and heart rate (HR) were monitored by telemetry in rats subjected to 1 h of physical restraint while Ta was maintained at 1430 degrees C in 2 degrees C increments. The Tc of unrestrained rats was unaffected by Ta. During restraint, Tc was elevated at ambient temperatures with the exception of 14 degrees C, at which the rats became mildly hypothermic. There was an inverse relationship between Ta and HR in both unrestrained and restrained rats; however, HR was significantly elevated in restrained rats at all ambient temperatures except 22 and 24 degrees C. Heat loss from the tail, estimated from Tc and tail skin temperature, was markedly reduced at all but the highest ambient temperatures in restrained rats. The data suggest that the Ta limits of normothermia are narrowed in the restrained rat. That is, between 16 and 20 degrees C, the rat maintains a relatively stable Tc that is slightly elevated above that of the unrestrained rat. At ambient temperatures above or below this range, the rat shows signs of hyperthermia and hypothermia, respectively. In contrast, the limits of normothermia for unrestrained rats range from 14 (or lower) to 30 degrees C. Overall, the ideal Ta for restrained rats appears to be 20 degrees C and no higher than 22 degrees C for the thermoregulatory system to maintain a regulated Tc in rats well adapted to physical restraint.