BACKGROUND Cardioplegia infusion pressure is usually not directly monitored during neonatal heart operations. We hypothesize that the immature newborn heart may be damaged by even moderate elevation of cardioplegic infusion pressure, which in the absence of direct aortic monitoring may occur without the surgeon's knowledge. METHODS Twenty neonatal piglets received cardiopulmonary bypass and the heart was protected for 70 minutes with multidose blood cardioplegia infused at an aortic root pressure of 30 to 50 mm Hg (low pressure) or 80 to 100 mm Hg (high pressure). Group 1 (n = 5, low pressure), and group 2 (n = 5, high pressure) were uninjured (nonhypoxic) hearts. Group 3 (n = 5, low pressure) and group 4 (n = 5, high pressure) first underwent 60 minutes of ventilator hypoxia (FiO2 8% to 10%) before initiating cardiopulmonary bypass to produce a clinically relevant hypoxic stress before cardiac arrest. Function was assessed using pressure volume loops (expressed as a percentage of control), and coronary vascular resistance was measured with each cardioplegic infusion. RESULTS In nonhypoxic (uninjured) hearts (groups 1 and 2) cardioplegic infusion pressure did not significantly affect systolic function (end systolic elastance, 104% versus 96%), preload recruitable stroke work (102% versus 96%) diastolic compliance (152% versus 156%), or coronary vascular resistance but did raise myocardial water (78.9% versus 80.1%; p < 0.01). Conversely, if the cardioplegic solution was infused at even a slightly higher pressure in hypoxic hearts (group 4), there was deterioration of systolic function (end systolic elastance, 28% versus 106%) (p < 0.001) and preload recruitable stroke work (31% versus 103%; p < 0.001), rise in diastolic stiffness (274% versus 153%; p < 0.001), greater myocardial edema (80.5% versus 79.6%), and marked increase in coronary vascular resistance (p < 0.001) compared to hypoxic hearts given cardioplegia at low infusion pressures (group 3), which preserved function. CONCLUSIONS Hypoxic neonatal hearts are very sensitive to cardioplegic infusion pressures, such that even moderate elevations cause significant damage resulting in myocardial depression and vascular dysfunction. This damage is avoided by using low infusion pressures. Because small differences in infusion pressure may be difficult to determine without a direct aortic measurement, we believe it is imperative that surgeons directly monitor cardioplegia infusion pressure, especially in cyanotic patients.