This study investigates whether the addition of magnesium to a hyperkalemic cardioplegic solution containing 0.1 mM ionized calcium improves myocardial preservation, and whether there is an optimal magnesium concentration in this solution. Isolated perfused rat hearts were arrested for two hours by this cardioplegic solution, which was fully oxygenated and infused at 8 degrees C every 15 minutes to simulate clinical conditions. The cardioplegic solution contained either 0, 2, 4, 8, 16, or 32 mM magnesium. At end-arrest, the myocardial creatine phosphate concentration (nanomoles per milligram of dry weight) was 20.7 +/- 2.1, 22.9 +/- 1.7, 24.8 +/- 2.0, 31.3 +/- 1.4, 33.1 +/- 1.8, and 31.6 +/- 0.8, respectively, in hearts given cardioplegic solution containing these magnesium concentrations. Thus, the concentration of creatine phosphate was significantly higher at end-arrest when the cardioplegic solution contained 8, 16, or 32 mM than 0 or 2 mM magnesium (p less than 0.002) or 4 mM magnesium (p less than 0.02), and highest with 16 mM magnesium. Also, creatine phosphate was more sensitive to the magnesium concentration of the cardioplegic solution than was end-arrest adenosine triphosphate levels, which did not differ among the experimental groups. Aortic flow, expressed as a percentage of prearrest aortic flow, was 60.3 +/- 5.0, 70.2 +/- 5.5, 71.6 +/- 4.4, 71.8 +/- 4.8, 81.0 +/- 5.0, and 71.8 +/- 5.3, respectively. The addition of magnesium to the cardioplegic solution improved recovery of aortic flow (p less than 0.05, 16 mM versus 0 mM magnesium). We conclude from these data that with deep myocardial hypothermia and at an ionized calcium concentration of 0.1 mM, the addition of magnesium, over a broad concentration range, improved preservation of myocardial creatine phosphate and, at a concentration of 16 mM, improved aortic flow. The optimal magnesium concentration in the cardioplegic solution was 16 mM.