Objective: Studies showing the effectiveness of therapeutic hypothermia (32-34 degrees C) in postcardiac arrest patients have been criticized because of patients with elevated body temperature (>37.5 degrees C) in the noncooled control group. Thus, the effects of spontaneous normothermia (<37.5 degrees C) compared with mild therapeutic hypothermia were studied. Design: Retrospective chart review from 1991 to 2010. Patients: Witnessed out-of-hospital arrest, presumed to be of cardiac origin, aged 18 to 80 yrs and with a Glassgow Coma Scale score of <8 at admission. Interventions: Patients with sustained restoration of spontaneous circulation who did not receive therapeutic hypothermia and never exceeded 37.5 degrees C during the 36 hrs postcardiac arrest were compared with patients who received mild therapeutic hypothermia. Measurements and Main Results: The primary end point was a favorable neurological outcome, defined as Cerebral Performance Categories 1 or 2; the secondary end point was overall survival to 180 days. Significantly more patients in the hypothermia group had Cerebral Performance Categories 1 or 2 (hypothermia: 256 of 467 [55%] vs. normothermia: 69 of 165 [42%]) and survived for >180 days (hypothermia: 315 of 467 [67%] vs. normothermia: 79 of 165 [48%]). The propensity score adjusted risk ratio for good neurological outcome of patients undergoing hypothermia treatment was 1.37 (confidence interval 1.09-1.72, p <= .01) and for dying within 180 days was 0.57 (confidence interval 0.44-0.73, p <= .01) compared to normothermia. Conclusions: Therapeutic hypothermia is associated with significantly improved neurological outcome and 180-day survival compared to spontaneous normothermia in cardiac-arrest patients. (Crit Care Med 2012; 40: 2315-2319)