STUDY OBJECTIVE To compare the rates of rewarming of forced-air and passive insulation as a treatment for accidental hypothermia. METHODS We carried out a prospective, randomized clinical trial in two urban, university-affiliated emergency departments. Our subjects were 16 adult hypothermia victims with core temperatures less than 32 degrees C. A convective cover inflated with air at about 43 degrees C (forced-air group) or cotton blankets (control group) were applied until the patient's core temperature reached 35 degrees C. Members of both groups were given IV fluids warmed to 38 degrees C and warmed, humidified oxygen at 40 degrees C by inhalation. RESULTS The mean +/- SD initial temperature was 28.8 degrees +/- 2.5 degrees C (range, 25.5 degrees C to 31.9 degrees C) in the patients who underwent forced-air rewarming and 29.8 degrees +/- 1.5 degrees C (range, 28.2 degrees C to 31.9 degrees C) in those given blankets. Core temperature increased about 1 degree C/hour faster in patients treated with forced-air rewarming (about 2.4 degrees C/hour) than in patients given only cotton blankets (about 1.4 degrees C/hour, P = .01). Core-temperature afterdrop was detected in neither group. CONCLUSION Forced air accelerated the rate of rewarming without producing apparent complications in hypothermic patients.