Fifty consecutive patients admitted within 12 hours of the onset of symptoms of acute myocardial infarction were randomly assigned to treatment with intravenous glucose-insulin-potassium infusion (23 patients) or to a control group (0.5 N sodium chloride infusion) (27 patients). The glucose-insulin-potassium infusion consisted of 30 g glucose, 50 U regular insulin and 80 mEq KCl per liter infused at 1.5 ml/kg per hour for 2 days. Serial measurements were made of pulmonary arterial end-diastolic pressure, cardiac index, daily fluid intake and output, serum glucose, potassium, urea nitrogen, free fatty acids, osmolarity, creatine kinase-MB isoenzyme and cardiac rhythm. Although all patients admitted comatose died (three glucose-insulin-potassium recipients, one control subject), hospital mortality in patients admitted noncomatose was 0 percent (0 of 20) in glucose-insulin-potassium recipients versus 12 percent (3 of 26) in the control group (three deaths secondary to late pump failure). Glucose-insulin-potassium recipients experienced 4.9 ± 1.3 hours of three or more premature ventricular complexes/min compared with 11.1 ± 1.9 hours for control subjects (P < 0.02). Twelve control patients had 60 episodes of ventricular tachycardia compared with seven glucose-insulin-potassium recipients, who had 12 episodes of ventricular tachycardia. During glucose-insulin-potassium infusion, fluid intake and output, serum glucose and potassium were significantly increased compared with values in control subjects, blood urea nitrogen and free fatty acids were decreased, whereas osmolarity, pulmonary arterial end-diastolic pressure and cardiac index did not differ significantly from control values. Creatine kinase-MB infarct size averaged 124 ± 15 IU/liter in glucose-insulinpotassium recipients and 109 ± 14 IU/liter in control subjects (difference not significant). These data demonstrate that, in patients with acute infarction, glucose-insulin-potassium infusion (1) does not adversely alter hemodynamics, (2) reduces free fatty acids, (3) diminishes frequency of ventricular arrhythmias, but (4) has no demonstrable effect on infarct size as assessed with creatine kinase isoenzyme values. In this ongoing randomized clinical trial, the number of patients studied is too small to permit definite conclusions to be reached regarding the effect of glucose-insulin-potassium infusion on hospital survival. © 1979.