Exercise testing alone or in combination with thallium scintigraphy has significant prognostic value. In contrast, dipyridamole thallium imaging is not dependent on patients achieving adequate levels of exercise, but no long-term prognostic studies have been reported. Accordingly, imaging results of 516 consecutive patients referred for dipyridamole thallium studies were correlated with subsequent cardiac events, death (n = 23) and myocardial infarction (n = 43) over a mean follow-up period of 21 months. Patients with a history of congestive heart failure, prior myocardial infarction, diabetes mellitus or abnormal scans were significantly more likely to have a cardiac event (p less than 0.03). With use of logistic regression analysis, an abnormal scan was an independent and significant predictor of subsequent myocardial infarction or cardiac death and increased the relative risk of any event more than threefold. The presence of redistribution on thallium scanning further increased the risk of a cardiac event. Survival analysis demonstrated a significant difference between patients with an abnormal or normal thallium scan over a 30 month period. In conclusion, dipyridamole thallium scintigraphy demonstrates prognostic value in a large unselected population and may be an adequate clinical alternative to physiologic exercise testing in the evaluation of coronary heart disease.