The hypothesis that hypothermic individuals can be actively rewarmed in the field by immersion of the extremities in hot water was investigated. Three techniques for rewarming subjects with lowered deep body temperatures were compared: a) whole body immersion to the neck in water at 40 degrees C; b) immersion of two hands plus forearms only in water at 42 degrees C; and c) passive rewarming. The suggestion that the fall in deep body temperature resulting from immersion to the neck in water at 15 degrees C could be arrested by immersing both arms in water at 42 degrees C was also investigated. Results indicated that immersion to the neck in hot water was clearly the most effective rewarming technique. No significant difference (p > 0.05) was observed in the deep body temperature response during passive rewarming or during immersion of both hands and forearms in water at 42 degrees C. In the later condition some increase in peripheral blood flow to the hands may have occurred and resulted in a heat input of approximately 12 W, but any benefit from this was negated by an associated significant decrease (p > 0.05) in intrinsic heat production. Immersing the arms in hot water during immersion to the neck in cold water appeared to accelerate rather than decelerate the rate of fall of deep body temperature. We concluded that hand rewarming, although theoretically attractive, is ineffective in practice and could be detrimental in some circumstances, by suppressing intrinsic heat production or precipitating rewarming collapse.