The assumption that an alkaline cardioplegic solution is advantageous to the arrested heart is not supported by experimental evidence. Three groups of dogs were subjected to 2 hours of hypothermic arrest using a potassium cardioplegic solution of varying pH: group 1 (n = 8), pH 7.1; group 2 (n = 6), pH 7.4; and group 3 (n = 8), pH 7.7. Left ventricular (LV) function curves were obtained before arrest and after reperfusion, and the percent recovery of LV function was calculated. Coronary blood flow (CBF), oxygen and lactate use, and adenosine triphosphate and creatine phosphate were measured before and after arrest. Recovery of LV function was 83 +/- 4% (mean +/- SEM) in group 1 and 75 +/- 6% in group 2. LV function in group 3 was significantly depressed at 51 +/- 5% (p less than 0.01). Immediately after arrest, CBF and oxygen consumption were lowest in group 1 and highest in group 3 (p less than 0.01). CBF increased 23 +/- 9% in group 1, 110 +/- 29% in group 2 and 277 +/- 33% in group 3. Oxygen consumption was 2.2 +/- 0.4 vol/100 g/min in group 1, 3.63 +/- 0.5 vol/100 g/min in group 2 and 4.63 +/- 0.7 vol/100 g/min in group 3. All groups showed postarrest depression of adenosine triphosphate and creatine phosphate, but return of creatine phosphate to control was slowest in group 1 (p less than 0.01). These results indicate no advantage to an alkaline cardioplegic solution and suggest that an acidic cardioplegic solution, by inhibiting metabolism, may increase protection of the arrested heart.