Most published studies of thermoregulatory responses of mice to LPS involved a stressful injection of LPS, were run at a poorly controlled and often subneutral ambient temperature (T(a)), and paid little attention to the dependence of the response on the LPS dose. These pitfalls have been overcome in the present study. Male C57BL/6 mice implanted with jugular vein catheters were kept in an environmental chamber at a tightly controlled T(a). The relationship between the T(a)s used and the thermoneutral zone of the mice was verified by measuring tail skin temperature, either by infrared thermography or thermocouple thermometry. Escherichia coli LPS in a wide dose range (10(0)-10(4) microg/kg) was administered through an extension of the jugular catheter from outside the chamber. The responses observed were dose dependent. At a neutral T(a), low (just suprathreshold) doses of LPS (10(0)-10(1) microg/kg) caused a monophasic fever. To a slightly higher dose (10(1.5) microg/kg), the mice responded with a biphasic fever. To even higher doses (10(1.75)-10(4) microg/kg), they responded with a polyphasic fever, of which three distinct phases were identified. The dose dependence and dynamics of LPS fever in the mouse appeared to be remarkably similar to those seen in the rat. However, the thermoregulatory response of mice to LPS in a subthermoneutral environment is remarkably different from that of rats. Although very high doses of LPS (10(4) microg/kg) did cause a late (latency, approximately 3 h) hypothermic response in mice, the typical early (latency, 10-30 min) hypothermic response seen in rats did not occur. The present investigation identifies experimental conditions to study LPS-induced mono-, bi-, and polyphasic fevers and late hypothermia in mice and provides detailed characteristics of these responses.