Intravenous nitroglycerin has beneficial effects on hemodynamics and on myocardial ischemia. The aim of the present study was to investigate whether it also reduces final infarct size. In 60 patients with myocardial infarction serial blood samples were tested for creatine kinase (CK) and its isoenzyme CK-MB activity for infarct size calculations. Hemodynamic parameters were measured every 4 hours. Patients were randomized to a control (n = 29) and a nitroglycerin group (n = 31). In the early intervention group continuous perfusion of nitroglycerin for 48 hours was started < 8 hours (mean 4.5 hours) after onset of symptoms, and in the late intervention group > 8 hours (mean 12.8 hours) after onset of symptoms. In early intervention (n = 22) peak CK activity was 871 U/l in control patients and 544 U/l (p < 0.05) in the nitroglycerin group. The rate of CK release was reduced from 79 to 33 U/l x h, i.e. by a total of 58%, as was total CK and CK-MB release (p < 0.02). Calculated CK infarct size was 69 g equiv. in controls and 48 g equiv. in the nitroglycerin group. (CK-MB: 68 versus 43 g equiv., p < 0.05). In late intervention (n = 28) the differences were less pronounced. Nitroglycerin reduced left ventricular filling pressure significantly and increased cardiac output. Blood pressure changed insignificantly and peripheral vascular resistance decreased. In conclusion, nitroglycerin reduces CK and CK-MB release and hence calculated infarct size, particularly when treatment is begun within the first 8 hours after onset of symptoms.