Although suppression of thermoregulatory mechanisms by anesthetics is generally assumed, the extent to which thermoregulation is active during general anesthesia is not known. The only thermoregulatory responses available to anesthetized, hypothermic patients are vasoconstriction and non-shivering thermogenesis. To test anesthetic effects on thermoregulation, the authors measured skin-surface temperature gradients (forearm temperature--finger-tip temperature) as an index of cutaneous vasoconstriction in unpremedicated patients anesthetized with 1% halothane and paralyzed with vecuronium during elective, donor nephrectomy. Patients were randomly assigned to undergo maximal warming (warm room, humidified respiratory gases, and warm intravenous fluids; n = 5) or standard temperature management (no special warming measures; n = 5). Skin-surface temperature gradients greater than or equal to 4 degrees C were prospectively defined as significant vasoconstriction. Normothermic patients [average minimum esophageal temperature = 36.4 +/- 0.3 degrees C (SD)] did not demonstrate significant vasoconstriction. However, each hypothermic patient displayed significant vasoconstriction at esophageal temperatures ranging from 34.0 to 34.8 degrees C (average temperature = 34.4 +/- 0.2 degrees C). These data indicate that active thermoregulation occurs during halothane anesthesia, but that it does not occur until core temperature is approximately equal to 2.5 degrees C lower than normal. In two additional hypothermic patients, increased skin-temperature gradients correlated with decreased perfusion as measured by a laser Doppler technique. Measuring skin-surface temperature gradients is a simple, non-invasive, and quantitative method of determining the thermoregulatory threshold during anesthesia.