Over the last few years there has been a renewal of interest in retrograde administration of cardioplegic solutions but this method has not been validated by a large scale clinical trial. From 1980 to 1989 our group used retrograde coronary sinus perfusion as the only means of administering cardioplegia in 500 consecutive patients undergoing isolated aortic valve replacement (AVR) (359 cases) or associated with another valvular or myocardial revascularization procedure (141 cases). Using this method, there were 31 hospital deaths (6.2%) of which 20 were of cardiac origin (4%). During the last 5 years, hospital mortality for isolated AVR decreased to 1.5%. With an incidence of low cardiac output of 12%, of arrhythmias and atrioventricular block of 7.4% and 1.2% respectively, and of myocardial infarction of 1.4%, this method of cardioplegia compares favourably with anterograde perfusion. There were only 3 traumatic lesions of the coronary sinus, all occurring at the beginning of the series and all of which were successfully repaired. Retrograde coronary sinus perfusion is therefore a safe and effective method of cardioplegia during aortic valve surgery: the cardioplegic solution is distributed uniformly in patients with coronary artery disease and therefore, this could become the technique of choice for myocardial protection in coronary artery or mixed coronary and valvular surgery.