The protection afforded by cardioplegia during elective ischemic arrest can be partly compromised by a reperfusion injury, which may impede the recovery of cardiac function. We previously showed experimentally that this postischemic damage could be largely avoided by an appropriate crystalloid reperfusate. The present study was thus undertaken to assess the effects of this 'reperfusion solution' clinically. One hundred twelve patients undergoing valve replacement with the aid of hypothermic cardioplegia (K+ 12 mEq, Mg2+ 26 mEq) were prospectively divided in two groups: Groups 1 (n = 49) received an unmodified blood reperfusate. In Group II (n = 63), 1 L of the reperfusion solution was delivered just prior to removal of the aortic clamp. The formulation of the reperfusion solution adhered to the following principles: (1) maintenance of cardioplegia (K+ = 15 mEq), (2) replenishment of Ca2+ stores (Ca2+ = 2.5 mEq), (3) substrate provision (glutamate = 2,942 gm), (4) buffering (pH = 7.70 at 28° C), and (5) hyperosmolarity (370 mOsm). The two groups were matched for preoperative data except for a higher incidence of isolated aortic valve replacement (p = 0.01) in Group II. Also, the cross-clamp time (mean ± standard error of the mean) was longer in Group II (94 ± 4 minutes versus 63 ± 4 minutes, p < 10-6). The reperfusion solution was found to increase both the rate and extent of postischemic functional recovery, as evidenced by (1) a lower proportion of catecholamine-supported patients 48 hours after operation (9/63 [14.28%] versus 16/49 [32.6%] in the control group [p < 0.03]) and (2) a lower amount (γ/kg/min) of dobutamine required to achieve stable hemodynamics (11 ± 1 versus 26 ± 6 in the control group [p < 0.03]). A similar recovery pattern was noted in the high-risk subgroup of patients with mitral valve disease. Further, serial postoperative hemodynamic measurements were performed in 31 randomly selected patients (10 control and 21 reperfused). Although the reperfused patients were found to be at higher risk because of lower preoperative cardiac indices and longer cross-clamp times, they consistently achieved better postoperative cardiac indices and longer cross-clamp times, they consistently achieved better postoperative hemodynamics with a lower incidence of catecholamine support. This hemodynamic improvement was particularly reflected by a higher left ventricular stroke work index throughout the postoperative course, the difference being significant 6 hours and 12 hours postoperatively. We conclude that the protection afforded by cardioplegia is significantly improved by an appropriate asanguineous initial reperfusate, especially in case of poor preoperative cardiac function and/or extended periods of ischemic arrest. The advantages of this crystalloid reperfusion solution over a modified blood reperfusate are discussed.