When systemic hypothermia was maintained before inducing cardiac arrest, the likelihood of successful defibrillation and meaningful survival was increased. When hypothermia is induced during cardiopulmonary resuscitation, mortality is also improved. With the introduction of the amplitude spectrum area as a predictor of the success of electrical defibrillation, we investigated the effect of preferential head cooling initiated coincident with cardiopulmonary resuscitation on amplitude spectrum area as a predictor. We hypothesized that rapid head cooling initiated coincident with cardiopulmonary resuscitation improves amplitude spectrum area, and therefore is predictive of successful defibrillation. Prospective randomized controlled study. University-affiliated research institute. Domestic pigs. Sixteen pigs, weighing 40.6 +/- 1.4 kg, were randomized to the hypothermia (n = 8), or control (n = 8) group. Ventricular fibrillation was induced and untreated for 10 mins. Cardiopulmonary resuscitation was then initiated for 5 mins followed by attempted defibrillation with a biphasic 150-J electric shock. Coincident with starting cardiopulmonary resuscitation, hypothermia was induced with evaporative intranasal cooling using a perfluorochemical. If spontaneous circulation was not restored after defibrillation, cardiopulmonary resuscitation was resumed for 1 min before the next defibrillation attempt until the animal was either successfully resuscitated or for a total of 15 mins. The target core temperature was 34 degrees C. Control animals were identically treated except for hypothermia. Five seconds of ventricular fibrillation waveform were recorded immediately preceding delivery of a shock. The ventricular fibrillation waveforms were analyzed using the amplitude spectrum area algorithm. A smaller epinephrine dose (60 +/- 32.1 vs. 30 +/- 0 mg/mL, p = .01) and shorter cardiopulmonary resuscitation duration (365 +/- 42 sec vs. 600 +/- 243 sec, p = .01) were required to achieve return of spontaneous circulation in the hypothermia group, compared with control. Five minutes after starting cardiopulmonary resuscitation, head temperature was reduced from 38 degrees C to 34 degrees C in the hypothermia group (p = .028). Hypothermia improved the success of electrical shocks before return of spontaneous circulation (88 +/- 18% vs. 66 +/- 19%, p = .034). Both the amplitude spectrum area values of initial shock (26.1 +/- 5.3 vs. 21.4 +/- 2.16 mV-Hz, p = .049) and total shocks (26.1 +/- 5.3 vs. 21.4 +/- 2.16 mV-Hz, p = .006) were significantly higher in the hypothermia group than control. Amplitude spectrum area served as a useful predictor for improved resuscitation and facilitated defibrillation in the setting of rapid head cooling initiated coincident with cardiopulmonary resuscitation.