Oxygenated Fluosol-43 cardioplegia (CP), a perfluorocarbon with high oxygen solubility, was compared with crystalloid and oxygenated blood cardioplegia. Potassium in each CP was 25 mEq/l. Thirty perfused rabbit hearts in three groups of 10 hearts each underwent 100 minutes of global ischemia at 20 degrees C, followed by 45 minutes of reperfusion at 37 degrees C. During ischemia, CP was given every 20 minutes. With each CP injection, increases in myocardial oxygen tension were recorded using mass spectrometry and oxygen consumption (MVO2) was calculated. Left ventricular function was assessed before and after ischemia by measuring isovolumic developed pressure and dP/dt with an intraventricular balloon. Intramyocardial PO2 increased by 19.6 +/- 1.8 mm Hg in the Fluosol CP group, 0.4 +/- 0.1 mm Hg in the crystalloid CP group and 1.5 +/- 0.3 mm Hg in the blood CP group (p less than 0.001, Fluosol CP vs crystalloid CP and blood CP). MVO2 with each CP injection, expressed as ml O2/100 g dry weight, was 203.8 +/- 7.0 for Fluosol CP, 20.4 +/- 1.2 for crystalloid CP and 39.2 +/- 4.3 for blood CP (p less than 0.001 Fluosol CP vs crystalloid CP and blood CP). Recovery of maximal dP/dt after 45 minutes of reperfusion, expressed as a percentage of preischemic control, was 75.6 +/- 4.0% for Fluosol CP, 60.9 +/- 5.5% for crystalloid CP and 53.4 +/- 3.7% for blood CP (p less than 0.02 Fluosol CP vs blood CP and crystalloid CP). These data clearly show that the use of Fluosol cardioplegic solution enhanced oxygen delivery and use compared with blood and crystalloid cardioplegic solutions. The marked increase in intramyocardial oxygen and MVO2 with each injection of Fluosol CP shows that there is effective aerobic metabolic activity during ischemia, which may explain the improved functional recovery. The failure of blood CP to afford similar protection can be explained by a decreased oxygen release from hemoglobin due to the leftward shift of the oxygen-hemoglobin dissociation curve with hypothermia.