Infusion of insulin-glucose-potassium is used to support the failing heart after cardiac operations. Although the effects on myocardial uptake of carbohydrates and lipids have been described, the effects on myocardial extraction of amino acids are unknown. This study was undertaken to clarify the effect of insulin-glucose-potassium on the pattern of amino acid uptake/release in myocardial and skeletal muscle after coronary operations. The amino acid uptake/release of the heart and of the leg was studied in 18 patients 1 hour after coronary bypass operations. The patients were randomized to treatment with 25 U of fast-acting insulin as a bolus injection followed by a continuous infusion of 1 U/kg body weight for 1 hour, or to serve as control patients. The hyperinsulinemic "clamp" technique was used to keep blood glucose unchanged during the study. In the insulin-treated group, the arterial concentration of 17 of 22 individual amino acids, including the three branched chain amino acids, decreased, the remainder being unchanged. The amino acid uptake/release of the leg was unchanged. The net myocardial uptake of leucine and isoleucine shifted to a no-uptake/no-release in the insulin-treated group, whereas the no-uptake/no-release of tyrosine and phenylalanine turned into a significant release. A positive correlation was observed between arterial concentration and myocardial uptake/release of the three branched chain amino acids. It is suggested that insulin, by lowering the arterial concentration of leucine and isoleucine, inhibited the myocardial uptake of these amino acids. This may have a negative effect on postoperative myocardial protein balance suggested by the release of tyrosine and phenylalanine.