In order to assess the prognostic significance of normal exercise thallium-210 myocardial scintigraphy in patients with documented coronary artery disease, we studied the incidence of cardiac death and non-fatal myocardial infarction in 69 symptomatic patients without prior Q wave myocardial infarction, who demonstrated one or more significant coronary lesions (stenosis > or = 70%) on an angiogram performed within 3 months of scintigraphy (Group 1). These patients were compared to a second group of 136 patients with an abnormal exercise scintigram, defined by the presence of reversible defect(s) and angiographically proven coronary artery disease (Group 2), and to a third group of 102 patients with normal exercise scintigraphy without significant coronary lesions (stenosis < or = 30%) or with normal coronary angiography (Group 3). In contrast to coronary lesions observed in Group 2, patients in Group 1 presented more frequently with single-vessel disease (83% vs 35%, P < 0.0001) and with more distal lesions (55% vs 23%, P < 0.0001). Over a mean follow-up period of 8.6 years, one fatal and eight non-fatal cases of myocardial infarction were observed in Group 1. The majority of patients in Group 1 were treated medically: only 24 (35%) underwent myocardial revascularization, usually by coronary angioplasty. There was no significant difference in the incidence of combined major cardiac events (cardiac death, non-fatal myocardial infarction) in patients with normal exercise scintigraphy, with or without documented coronary artery disease (Groups 1 and 3), while the incidence was higher in Group 2. However, while the mortality remained very low in Group 1, the incidence of non-fatal myocardial infarction was not different from that of Group 2, where most patients underwent revascularization procedures. In conclusion, patients with coronary artery disease and a normal exercise thallium-201 myocardial scintigram usually have mild coronary lesions (single-vessel disease, distal location) and good long-term prognosis, with a low incidence of cardiac death.