During two hours of aortic clamping, two groups of 10 dogs each were given an intermittent infusion of a cold solution in the aortic root. In one group, the solution contained 20 mEq per liter of potassium chloride (KCl) and in the other, a calcium channel blocker (diltiazem). Left ventricular (LV) performance was measured by calculation of LV pressure, left ventricular end-diastolic pressure (LVEDP), cardiac index (CI), and stroke-work index (SWI). Regional myocardial function was assessed through ultrasonic crystals implanted in the subendocardial areas of both the left anterior descending (LAD) and circumflex coronary arteries. Dogs receiving KCl displayed deterioration of LV performance as evidenced by a return of maximal LV pressure, maximal rate of rise of LV pressure (dP/dtmax), CI, and SWI to 74 +/- 4%, 87 +/- 5%, 74 +/- 6%, and 59 +/- 6%, respectively, of the initial (before clamping) values. Animals that received diltiazem, on the other hand, had for the same variables a return to 85 +/- 4%, 99 +/- 7%, 129 +/- 8%, and 111 +/- 10% of the initial values. The rate of relaxation (peak negative dP/dt) decreased in both groups but less in dogs receiving diltiazem. Regional function in the area of the LAD and circumflex arteries showed little change in either group. We conclude that cold cardioplegia with a solution containing KCl or diltiazem protects the myocardium during prolonged ischemic cardiac arrest. Return of LV function on the whole is superior when diltiazem is used.