Oxygen free radicals (OFRs) are associated with ischaemia-reperfusion injury involving many organs, including the heart, which can lead to depressed cardiac function and abnormalities in the cardiac ultrastructure. This is seen upon the release of the aortic crossclamp when the ischaemic myocardium is reperfused in patients undergoing cardiopulmonary bypass (CPB). Various studies have shown that by adding OFR scavenging agents or antioxidants to the CPB prime or cardioplegia, cardiac performance improves. Mannitol is an osmotic diuretic with free radical scavenging properties, which has been shown to reduce the extent of ischaemic injury and improve the function of the myocardium. This study evaluated how effective mannitol is as an OFR scavenger by administering different concentrations of cardioplegia antegrade into the aortic root, thus maximising its effects directly upon the myocardium rather than being diluted in the CPB prime. Thirty-three patients undergoing primary coronary artery bypass grafting (CABG) were, by double blind random selection, allocated into one of three groups: group 1, a control group (consisting of 11 patients) receiving no mannitol; group 2 (11 patients), receiving a concentration of 4 g/l; and group 3 (11 patients), receiving 8 g/l. Three blood samples were taken directly from the coronary sinus during bypass: the first sample at the start of bypass, just prior to the crossclamp being applied; the second sample just after removal of the crossclamp; and the third sample just prior to termination of bypass. All samples were then centrifuged and the plasma analysed for malondialdehyde (MDA) using high-performance liquid chromatography (HPLC). MDA, an endproduct of lipid peroxidation, causes cellular damage and disruption of cell membranes when tissue antioxidants are exhausted. The more MDA produced, the greater the depletion of tissue antioxidants secondary to OR formation during reperfusion when the aortic crossclamp is removed. HPLC is a useful biochemical study; however, it is not a direct indicator of depressed myocardial function, such as an invasive test would be, and this should be borne in mind. Statistically, the results do not show a significant difference among the three groups or among the three samples. However, a trend can be seen, which shows lower levels of MDA in the two groups receiving mannitol and there is an indication of a rise in MDA levels upon the start of reperfusion in the two groups receiving mannitol, but not the control group. It is concluded that further samples would be needed to find a significant difference in MDA concentrations.