The role of oxygen (O2) in blood cardioplegia (BCP) remains controversial. On the one hand, O2 reduces ischemic injury between BCP infusions by maintaining energy production through oxidative pathways. On the other hand, O2 carried by blood may not be released to the tissue at 4 degrees C or potentially provides substrate for deleterious O2 radical species. This study tests the hypothesis that O2 is a critical component in myocardial protection afforded by BCP. In 17 anesthetized dogs, left ventricular performance was measured by left ventricular end-systolic pressure-volume relations using the position of the end-systolic pressure-volume relation quantitated by the left ventricular midrange volume intercept at 100 mm Hg (V100) to describe performance. After 30 minutes of global normothermic ischemia, hearts were protected with multidose 4 degrees C BCP for 1 hour of arrest. Oxygen content in BCP was adjusted to 1.1 +/- 0.2 vol% (n = 7; desaturated BCP group), 4.3 +/- 0.5 vol% (n = 5; intermediate oxygenated BCP group), or 10.2 +/- 0.6 vol% (n = 5; saturated BCP group) using a membrane oxygenator interposed in the BCP circuit and aerated with an appropriate mixture of O2, nitrogen, and carbon dioxide. After 1 hour of 37 degrees C reperfusion, 3 of the 7 dogs in the desaturated BCP group failed to generate sufficient cardiac output to discontinue bypass. In the remaining 4 dogs, severe left ventricular depression caused a rightward shift in V100 from 17 +/- 4 to 47 +/- 9 mL (p = 0.02). With intermediate BCP, all hearts were weaned from bypass with marginal left ventricular depression (V100, 20 +/- 5 versus 46 +/- 16 mL; p = 0.10). In contrast, hearts protected with saturated BCP showed no significant increase in V100 (13 +/- 4 versus 24 +/- 13 mL; p = 0.23). We conclude that O2 in BCP is critical to its myocardial protective properties.