In two groups of animals (6 and 9 dogs), the aorta was cross-clamped 60 and 90 minutes, respectively, during hypothermic cardiopulmonary bypass. Immediately after cross-clamping, pharmacological cardiac arrest was induced by injecting 100 ml of a cold cardioplegic solution into the aortic root. Topical cardiac hypothermia was added. In hearts undergoing 90 minutes of ischemia, a repeat injection of the cardioplegic solution was done at 45 minutes. In 14 dogs (control group), only topical cardiac hypothermia was instituted for myocardial protection during 60 minutes of ischemia. Seven weeks after operation the surviving animals (6 in each group) were killed. Study of myocardial performance failed to demonstrate significant differences among the groups. Microscopic examination of transmural samples taken from anatomically defined sides of both ventricles, disclosed isolated, punctuate subendocardial scars in only 2 hearts of the control group. All the hearts having 90 minutes of pharmacological cardiac arrest and topical cardiac hypothermia exhibited diffuse fibrosis replacing 10 to 20% of the left ventricular myocardium. Extent and incidence of fibrosis were significantly higher in these hearts in comparison to those of the other groups. We conclude that pharmacological cardiac arrest plus topical cardiac hypothermia makes a safe and efficient method of myocardial protection during aortic cross-clamping only if the ischemic interval is limited to 60 minutes. It cannot prevent permanent myocardial injury if the ischemic arrest is extended to 90 minutes.