Glyceryl trinitrate was previously said to be contraindicated in patients with acute myocardial infarction. Its intravenous administration during acute infarction, however, was associated with a beneficial effect as determined by ST segment mapping. Most recently in a selected group of patients with acute infarction and abnormal haemodynamics, intravenous glyceryl trinitrate was shown to reduce infarct size estimated by enzymes. The present study was performed to verify the safety of intravenous glyceryl trinitrate in patients with infarction under conventional clinical conditions without invasive monitoring and to determine its effect on infarct size in a prospective randomised trial involving 85 patients with infarction (43 treated and 42 control). Treated patients received glyceryl trinitrate within 10 hours of the onset of symptoms (mean 6.0 hours), and the dose was titrated to preset limits for changes in heart rate and blood pressure. In patients with inferior infarction, infarct size estimated by enzymes in the treated was only 12.2 +/- 1.8 versus 19.1 +/- 3.6 CK gram equivalents per metre squared in the placebo group. A similar but statistically insignificant trend was observed for subendocardial infarction but no difference was observed for anterior infarction. Ventricular arrhythmias determined from 24 hour tapes were more frequent in treated patients though this was not statistically significant. Lignocaine requirements in treated and control (1692 +/- 250 vs 1512 +/- 232 mg/24 h) were similar, as were the requirements for morphine (11.4 +/- 1.8 vs 12.2 +/- 2.2 mg/24 h). Results indicate that intravenous glyceryl trinitrate can be administered safely during evolving infarction without invasive monitoring and reduces infarct size in patients with inferior infarction.