Exercise radionuclide ventriculography (RVG) was performed 7-12 days after acute myocardial infarction (MI) in 153 patients to assess its value in identifying those at risk of serious recurrent cardiac events. In addition to electrocardiogram (ECG) features of the exercise test, clinical and hemodynamic features were also considered: the exercise test was abnormal if there was a fall in blood pressure of 10 mmHg or more, development of angina, or inability to complete three minutes of exercise. RVG was used to measure left ventricular ejection fraction and to assess wall motion at rest and at peak exercise. After a median follow-up of 14 months, there were 18 cardiac events: six deaths and 12 patients with recurrent MI. In addition, 18 patients underwent coronary artery bypass surgery; the decision to perform surgery was predicted by ST segment depression or a fall in blood pressure during exercise (p less than 0.005). The sensitivity of the exercise test for identifying patients with a cardiac event increased progressively as additional parameters were considered in a cumulative fashion; hence, ST depression identified 28% of patients with events, whereas addition of clinical and hemodynamic parameters increased this to 61%. The addition of RVG increased the sensitivity further to 88%, but with a specificity of 50%. Although exercise-induced ST depression of up to 1 mm did not predict outcome significantly, a positive exercise RVG did predict cardiac events: 17% of patients with a positive test had an event, compared with 6% of those with a normal study (p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)