Efficacy of three different solutions for myocardial protection was determined in dogs after 120 minute ischemia. In group I cold cardioplegia and topical hypothermia were performed with Ringer's solution in 9 dogs. Cold Ringer's solution, which was modified to contain potassium at 30 mEq/1 and sorbitol at 401 mOsm/1, was used in group Ⅱ (6 dogs). In 11 dogs (group Ⅲ) Collins solution (K: 115 mEq/1, 467 mOsm) was employed. The parameters measured included myocardial cyclic AMP, myocardial adenocine triphosphate, blood lactate, blood pyrvate, blood glucose, blood NEFA and O2 consumption. Results obtained were as follows: 1. Resuscitation rate was highest in group Ⅱ, being 100%, and lowest in group I, 11%. It was 45% in group Ⅲ. 2. Myocardial adenosine triphosphate decreased least in group Ⅱ (3.83μmol/g) and most in group I (2.18μmol/g). 3. There were no statistically significant inter-group differences in the other parameters. It was concluded that the best solution for myocardial protection was the modified Ringer's solution in group Ⅱ because the resuscitation rate was the best of the 3 groups. It was also noted that the resuscitation rate coincided well with the concentration of ATP in the myocardium. The Collins solution was suitable for stopping cardiac action, but not suitable for coronary perfusate because of a high potassium concentration which injured the myocardium. Cold Ringer's solution was effective in myocardial protection only by cooling the heart, not by its own protective action.