Total and individual plasma free fatty acids (FFA) were measured on admission and over the next 4 days in 24 patients admitted to the hospital with chest pain and suspected acute myocardial infarction (AMI). In a prospective randomized fashion, the patients were either given an infusion of 300 gm of glucose, 50 units of inuslin, and 80 mEq of KCl per liter at a rate of 1.5 ml/kg/hr over the initial 48 hours of hospitalization, or they served as controls receiving conventional therapy. Eleven patients were in the control group and 13 were in the glucose-insulin-potassium (G-I-K) group. Twenty-one of the patients had an AMI by CK-MB rise and ECG changes (in the G-I-K group three did not evolve AMI). The total plasma FFA were 840 ± 134 μM/L in the controls and 933 ± 160 μM/L in the G-I-K group initially (prestudy). Total FFA rapidly fell in the G-I-K group and then rebounded when G-I-K was stopped. In contrast, total FFA values fell gradually in the control group over the 4-day period. The individual FFA had similar percentages initially in the two groups. In the control group the percent of individual plasma FFA was unchanged over the period studied, although there was some mild random day-to-day fluctuation. In contrast in the G-I-K group linoleic acid fell both during and after the infusion was stopped (26.8% to 19.1% P < 0.001). Arachidonic acid doubled in percentage of the total FFA value during G-I-K infusion (3.1% to 6.5%, P < 0.002) and returned to the control value when it was stopped. Thus G-I-K infusion during AMI reduces the total level of plasma FFA while increasing the percent of arachidonic and decreasing the percent of linoleic acid, observations proposed to reflect improved membrane stability of the ischemic myocardium. © 1981.