Thirty patients with acute myocardial infarction admitted 2.1 +/- 1.1 (mean +/- standard deviation) hours after the onset of pain and with S-T segment elevation in multiple leads in the standard electrocardiogram were given either intravenous morphine (15 patients) or sublingual nitroglycerin (15 patients), and the effect on pain and QRS changes was observed. Nitroglycerin was administered repetitively in large doses while systolic blood pressure was maintained above 100 mm Hg. Chest pain failed to respond within 30 minutes in two patients who received nitroglycerin. In the remaining 13 patients nitroglycerin produced partial relief of pain in 17 +/- 5 minutes and complete relief in 127 +/- 65 minutes, requiring a cumulative dosage of 23.7 +/- 38.7 mg in 16 +/- 7 divided doses. An average of 14.9 +/- 7.1 mg of morphine in 3.3 +/- 1.5 divided doses produced complete relief of pain in a similar period (134 +/- 77 minutes [difference not significant]). In patients receiving morphine, Q waves developed at 24 and 48 hours, respectively, in 62 (72 percent) and 66 (77 percent) of a total of 86 sites with initial S-T segment elevation in the standard 12 lead electrocardiogram. In nitroglycerin responders, Q waves developed at 24 and 48 hours, respectively, in only 21 (28 percent, p less than 0.001) and 22 (29 percent, p less than 0.001) of the 76 sites with initial S-T segment elevation. Other electrocardiographic estimates of the extent of myocardial necrosis, including the percent reduction in R wave amplitude and the relative changes in R and Q wave amplitude, also were significantly less in those receiving nitroglycerin. There was no in-hospital mortality. Thus, large and frequent doses of nitroglycerin when used in the hyperacute phase of acute myocardial infarction can effectively abolish chest pain and limit later electrocardiographic signs of myocardial necrosis.