The study hypothesis was that obliteration of the posterior interventricular vein in the coronary sinus avoids the back leak of cardioplegia to the right atrium and forces cardioplegia towards the posterior wall of the left ventricle and interventricular septum. A new retrograde cardioplegia cannula with a long balloon (3 cm) was designed which obstructs the posterior interventricular vein in the coronary sinus. The hypothesis was tested by a prospective randomized study in 52 consecutive patients who underwent coronary or aortic valve surgery. In group I (n = 26), the cannula prototype was used, while a standard cannula (balloon length 8 mm) was used in group II (n = 26). The cardioplegic solution was cold blood (14 degrees C). The posterior wall temperature was recorded when the anterior wall temperature reached 15 degrees C. In group I, 91% of patients had the same temperature in the anterior and posterior walls of the left ventricle versus 19% in group II (P < 0.05). The mean of the difference of left ventricular temperatures between anterior and posterior walls was 0.5 degrees C (sigma = 1.7) in group I versus 8 degrees C (sigma= 4.1 ) in group II (alpha < 0.05). In group I, 9.5% of patients had a posterior wall temperature > 20 degrees C versus 81% in group II (P < 0.05). Cooling of the posterior wall of the left ventricle is better in group I than in group II. As cooling and cardioplegia flow are closely linked, obliteration of the posterior interventricular vein in the coronary sinus improves left ventricular distribution of the cardiplegia.