The prognostic value of early clinical history, exercise testing and ambulatory electrocardiography was assessed in 263 men (mean age 50 years) recovering from an uncomplicated myocardial infarction (MI). During a mean follow-up period of 31 months, 11 patients died of cardiac causes, 22 developed a non fatal recurrent MI, 16 unstable angina (UA) and 16 underwent coronary artery bypass surgery. The appearance at the exercise stress test of an ischemic S-T segment depression of 0.2 mV or greater (P less than 0.001) as well as the achievement of a work load of 360 Kg-m/m' or less (P less than 0.01) and of a rate-pressure product of 200 Units or less (P less than 0.01), were found to be predictive of the future development of UA, but neither of cardiac death nor of non fatal recurrent MI. The ischemic response was also seen to be predictive of cardiac death (P less than 0.05). S-T segment depression of 0.1 mV or greater, angina and ventricular ectopic activity during the stress test and clinical history were not of predictive value. Complex ventricular ectopic activity (multiform extrasystoles, couplets and ventricular tachycardia) recorded during 24 hour ambulatory electrocardiogram was seen to be predictive of death and non fatal MI. Whereas some parameters such as the ejection fraction and the extension of coronary artery disease are generally accepted as good predictors for cardiac events, others, such as those derived from exercise testing, history and ambulatory electrocardiography may change their predictive value from one survey to another. These discrepancies are due to differences in patient characteristics, in methodology and in medical management.(ABSTRACT TRUNCATED AT 250 WORDS)