Cold cardioplegia has been appreciated as a superior method of myocardial protection, but the ideal coronary perfusate has not yet been established. In this study three types of solutions were tested as a coronary perfusate to choose the most suitable solution by comparing hemodynamics, enzymes and histology before and after clamping the aorta for 2 hours. In 6 canine hearts that were perfused with cold Ringer's solution (Group I), only one heart was able to be resuscitated and serum enzymes (GOT, LDH, CPK and CPK-MB) were prominently increased. Interstitial edema of the myocardium was severe histologically. In 9 hearts that were perfused with cold Collins' solution (Group II), 5 hearts were able to be resuscitated. The enzymes were all increased and myocardial interstitial edema was prominent. In 6 hearts that were arrested with cold Collins' solution and immediately thereafter perfused with cold Ringer's solution which contained 30 mEq/l of potassium (Group III), resuscitation was successful in all hearts. CPK-MB was not significantly increased and interstitial edema was mild. In Group I, myocardial dysfunction was thought to be due to anoxic injury. In Group II, the dysfunction was thought to be secondary to the excessively high concentration of potassium. In Group III, it was concluded that anoxic myocardial injury was prevented because cardiac arrest was maintained with 30 mEq/1 potassium and the injury due to the perfusate itself was also minimized because potassium concentration was lowered to 30 mEq/l and that the induced myocardial dysfunction was minimal.