Exercise thallium scintigraphy is widely used to assess prognosis in patients with suspected or proven coronary artery disease. The incremental prognostic value of this technique in patients who have good exercise tolerance has not been well studied. Two hundred ninety-nine patients with known or suspected coronary artery disease without prior myocardial infarction or revascularization procedure referred for exercise myocardial perfusion imaging and able to exercise to greater than or equal to stage III of the Bruce protocol were included. After a mean follow-up of 50 +/- 10 months, there were 15 cardiac events (5%). The incidence of cardiac events was 10 versus 3% (p less than 0.001) in patients with an abnormal versus normal thallium-201 scan, and 9 versus 3% (p = 0.03) for an abnormal versus normal exercise electrocardiogram. When the 185 patients with a normal exercise electrocardiogram were examined, the incidence of cardiac events was 3% (5 of 150) in patients with a normal scan versus 0% (0 of 35) in patients with an abnormal scan. In the 114 patients with an abnormal exercise electrocardiogram, an abnormal thallium-201 scan was predictive of cardiac events (18% [8 of 44] versus 3% [2 of 70]; p = 0.006). Stepwise logistic regression analysis selected an abnormal thallium-201 scan and abnormal exercise electrocardiogram, low peak exercise heart rate, and male gender as independent variables associated with a significant increased risk of cardiac events. Thus, in patients with known or suspected coronary artery disease and good exercise tolerance, the addition of thallium-201 imaging in patients with an abnormal exercise electrocardiogram provides useful prognostic information.(ABSTRACT TRUNCATED AT 250 WORDS)