BACKGROUND This study tests the hypothesis that continuous normothermic retrograde blood cardioplegia is superior to cold intermittent blood cardioplegia in protecting the left and right side of the heart transmurally during an extended cross-clamping period. METHODS Twelve anesthetized, open chest dogs were placed on cardiopulmonary bypass and randomized to receive continuous warm (n = 6) or intermittent cold cardioprotection (n = 6) during a 3-hour aortic cross-clamp period. Transmural left ventricular muscle biopsy specimens were taken before the initiation of cardiopulmonary bypass and 90 and 180 minutes after cross-clamping. Right ventricular (RV) biopsy specimens were taken 180 minutes after aortic cross-clamping. Biopsy specimens were analyzed for adenosine triphosphate, creatine phosphate, and lactate levels and for morphologic changes via electron microscopy. RESULTS At the end of 180 minutes of cardiopulmonary bypass, the adenosine triphosphate contents of endocardial and epicardial halves of the left ventricular myocardium were only slightly degraded in both cardioplegia groups; a significantly greater reduction in adenosine triphosphate levels occurred in the RV of the warm compared with the cold group (p < 0.02). The difference in creatine phosphate values in the left ventricle between the cold group (35.2 +/- 23.4 nmol/mg cardiac protein) and the warm animals (64.4 +/- 24.9 nmol/mg cardiac protein) was not statistically significant, but the RV creatine phosphate stores were significantly better preserved in the warm compared with the cold cardioplegia group (p < 0.02). Lactate levels increased to a similar extent in both groups, but both values rose significantly over baseline (p < 0.03). Importantly the electron microscopic score of the left ventricle and RV indicated that cells were reversibly and not irreversibly damaged with both cardioplegic protections. CONCLUSIONS These results suggest the following: (1) Chemical arrest is a major contributor of myocardial preservation during diastolic arrest as used in clinical cardiac surgery. (2) Both methods preserve the ultrastructure of the myocytes transmurally during 3 hours of aortic cross-clamping. (3) Both techniques protect the RV and left ventricle; however, to provide optimal protection of the RV, alternated retrograde and antegrade perfusion might be beneficial over retrograde cardioplegia flow alone, in particular with warm cardioplegia.