Recent experimental studies have demonstrated that mild hypothermia at about 34 degrees C can be effective in the control of intracranial hypertension. A randomized controlled study of mild hypothermia was carried out in 33 severely head-injured patients. All patients fulfilled the following criteria: 1) persistent intracranial pressure (ICP) greater than 20 mm Hg despite fluid restriction, hyperventilation, and high-dose barbiturate therapy; 2) an ICP lower than the mean arterial blood pressure; and 3) a Glasgow Coma Scale score of 8 or less. The patients were divided into two groups: one received mild hypothermia (16 patients) and one served as a control group (17 patients). Mild hypothermia significantly reduced the ICP and increased the cerebral perfusion pressure. Eight patients (50%) in the hypothermia group and three (18%) in the control group survived (p < 0.05), while five (31%) in the hypothermia group and 12 (71%) in the control group died of uncontrollable intracranial hypertension (p < 0.05). In five patients in the hypothermia group, cerebral blood flow was measured by the hydrogen clearance method and arteriojugular venous oxygen difference was evaluated before and during mild hypothermia. Mild hypothermia significantly decreased the cerebral blood flow, arteriojugular venous oxygen difference, and cerebral metabolic rate of oxygen (p < 0.01). The results of this preliminary investigation suggest that mild hypothermia is a safe and effective method to control traumatic intracranial hypertension and to improve mortality and morbidity rates.