Reperfusion of warm blood cardioplegic solution is useful in minimizing reperfusion damage after ischemia. This study tests the hypothesis that overzealous administration of blood cardioplegic solution at reperfusion counteracts these benefits and can lead to a prevalence of depressed ventricular performance and mortality similar to that seen after normal blood reperfusion. Thirty-one dogs underwent 45 minutes of 37 degrees C global ischemia on vented bypass. Six received normal blood reperfusion and 25 were reperfused with a warm aspartate/glutamate-enriched blood cardioplegic solution; of these, eight received high-dose (3600 +/- 600 ml) and 17 received limited-dose (1180 +/- 120 ml) blood cardioplegic reperfusion over 10 to 20 minutes. High-dose blood cardioplegic perfusion (5100 +/- 200 ml) without prior ischemia was tested in an additional five dogs. High-dose blood cardioplegia without preceding ischemia did not alter ventricular function (peak stroke work index 96% of control). After ischemia, normal blood reperfusion (no cardioplegia) resulted in marked left ventricular dysfunction (peak stroke work index 36% of control, p less than 0.05 versus control) and a 33% mortality rate (2/6 died). High-dose cardioplegic reperfusion yielded marginal recovery of stroke work index (40% of control, p less than 0.05 versus control) and a 25% mortality rate (2/8 died). In contrast, limited-dose reperfusion of blood cardioplegic solution allowed 100% survival (17/17) and restored stroke work index to 90% of control (1.3 versus 1.45 gm.m/kg). We conclude that reperfusion damage can be avoided by initial reoxygenation with limited doses of substrate-enriched blood cardioplegic solution. Conversely, high-dose reperfusion of blood cardioplegic solution offsets this benefit, reduces recovery substantially, and may be lethal.