Perioperative hypothermia usually results largely from pharmacologic inhibition of normal thermoregulatory control. Midazolam is a commonly used sedative and anesthetic adjuvant whose thermoregulatory effects are unknown. We therefore tested the hypothesis that midazolam administration impairs thermoregulatory control. Eight volunteers were studied on 2 days each, once without drug and once at a target total plasma midazolam concentration of 0.3 micrograms/mL (corresponding to administration of approximately 40 mg over approximately 4 h). Each day, skin and core temperatures were increased sufficiently to provoke sweating, and then reduced to elicit peripheral vasoconstriction and shivering. We mathematically compensated for changes in skin temperature using the established linear cutaneous contributions to control of each response. From these calculated thresholds (core temperatures triggering responses at a designated skin temperature of 34 degrees C), we determined the thermoregulatory effects of midazolam. The sweating threshold was decreased approximately 0.3 degrees C by midazolam administration: 37.3 +/- 0.2 degrees C vs 37.0 +/- 0.3 degrees C (P = 0.0004, paired t-test). Midazolam decreased the core temperature that triggered vasoconstriction somewhat more: 37.1 +/- 0.2 degrees C vs 36.3 +/- 0.5 degrees C (P = 0.0002). Similarly, midazolam decreased the shivering threshold: 35.9 +/- 0.3 degrees C vs 35.3 +/- 0.6 degrees C (P = 0.03). The sweating-to-vasoconstriction (interthreshold) range, therefore, increased from 0.2 +/- 0.1 degrees C to 0.7 +/- 0.3 degrees C (P = 0.002). Although statistically significant, this relatively small increase contrasts markedly with the 3-5 degrees C interthreshold ranges produced by clinical doses of volatile anesthetics, propofol, and opioids.(ABSTRACT TRUNCATED AT 250 WORDS)