We compared the speed of cooling and treatment efficacy for evaporative cooling vs. iced gastric lavage in a canine heatstroke model. Nine random-source, mongrel dogs were anesthetized, shaved, and internally heated until the core temperature reached 43.0 degrees C. The animals were then randomly assigned to be cooled to 37 degrees C either by iced (1 degree C) tap water gastric lavage (n = 5200 ml/min) through a large (32-Fr) orogastric tube, or by spraying with tap water (n = 4, 15 degrees C, 12 L/min) before a large fan blowing room temperature air (23 degrees C) across the dog at 0.5 m/sec from a height of 50 cm. Temperatures were monitored by thermocouples in both tympanic membranes and the pulmonary artery. BP, pulse, and cardiac output were measured every 5 min. Evaporative cooling was twice as fast as iced gastric lavage (0.16 +/- .05 degree vs. 0.08 +/- .01 degree C/min X m2, p less than .01). Animals in the evaporatively cooled group also experienced a quicker and more complete return to baseline cardiac indices than the lavage-treated group. Moreover, all animals treated with evaporation survived and were neurologically intact 48 h later, while only one lavage-treated dog was neurologically intact over the same period. The others in the lavage group died one hour after cooling (n = 1), were grossly ataxic (n = 1), or were persistently comatose (n = 2). A simple evaporative cooling technique, readily available in the emergency department, appears to be the most rapid and effective means for cooling and treating heatstroke in the dog.