Heat stroke and other hyperthermia-related crises are serious clinical problems in childhood and adolescence. Rapid cooling is required to reduce morbidity and mortality. A variety of effective cooling methods exist, and some may interfere with monitoring and resuscitation or are not readily available. We studied in 12 pigtail monkeys the pathophysiology of immersion hyperthermia (42 degrees C) to cardiac arrest (1 min no flow) and CPR plus cooling to normothermia for restoration and stabilization of spontaneous normotension. This was followed by intractable shock and secondary arrest. These studies gave us the opportunity to compare two simple cooling methods applied during and after CPR: group I (n = 6) received application of ice bags to the groins, axillae, and neck. Group II (n = 6) received ice bags plus cold water wetting (sponging) over the entire anterior surface of trunk and extremities, plus fanning. CPR restored spontaneous circulation in four of six in each group, after CPR of 1.5-16 min (NS between groups). Speed of cooling correlated with speed of stabilization of spontaneous normotension. After cardiac arrest and during and after CPR, rectal temperature had declined from a lethal level of 42.2 degrees C to a safe level of 38.5 degrees C within 45 +/- 6 (38-53) min in group I, and within 28 +/- 4 (23-32) min in group II (p less than 0.05). Epidural and esophageal temperatures declined more rapidly than rectal temperature. For critical hyperthermia, we recommend immediate application of ice bags, cold water wetting (sponging), fanning, and head cooling combined when invasive blood cooling (the most effective method) is not immediately available.