A study was undertaken to compare warm gastric (Group 1) and closed thoracic cavity (Group 2) lavage for rewarming severely hypothermic dogs. Adult mongrel dogs were monitored by intra-arterial catheter, central venous catheter, and ECG, and by central venous, esophageal, and rectal temperature probes. Animals were externally cooled to an average of 21.2 C using ice bags. Eight Group 1 and eight Group 2 animals underwent continuous warm saline gastric or closed thoracic cavity lavage, respectively, using afferent and efferent nasogastric and thoracostomy tubes. No animal suffered ventricular fibrillation during tube placement. The closed lavage system consisted of a high-efficiency heat exchanger, a roller pump infusion device, and a heat exchange fluid bath. The lavage fluid circulated at a flow rate of 550 mL/min and a temperature of 39 C. Thoracic lavage animals were followed clinically for 24 hours for evidence of complications, then euthanized and autopsied. The mean time required to rewarm the animals 10 C by central venous temperature probe was 210.9 +/- 18.6 minutes for the gastric group and 99.3 +/- 23.0 minutes for the thoracic group (P less than .001). Rectal temperature consistently lagged behind central venous temperature during both the cooling and rewarming phases in both treatment groups. All of the thoracic lavage animals made an uneventful recovery. Continuous warm saline thoracic cavity lavage for core rewarming of severely hypothermic dogs is more effective than gastric lavage, and appears to be safe.(ABSTRACT TRUNCATED AT 250 WORDS)