Coronary artery ligations were performed in anesthetized, closed-chested dogs using snares. Cardiac output was measured with an electromagnetic flowmeter. Hyperbaric oxygenation (HBO) at 2 atmabs was used and blood exchanged for a dextran solution resulting in a hemoglobin level of about 8 g/100 ml. Control groups were used to study the effect of the preparation itself and hemodilution or HBO treatment alone. Cardiac output was decreased by the preparation and further decreased by the ligations. Six of seven dogs treated with HBO and hemodilution survived versus three of eight comparable preparations in the ligation control group. Cardiac output was improved while the elevated left atrial pressure seen in the control group was not improved. All deaths were due to ventricular fibrillation except the single death in the HBO with hemodilution treatment group which was due to electrical-mechanical dissociation. The probability that this increased survival was due to chance was less than 10% (P < 0.10).