In a follow-up study (mean = 3.6 years) of 555 men under 40 years of age (mean = 35.7 years) who had recently suffered transmural myocardial infarction, 44 patients (7.9%) suffered cardiac death with an annual rate of 2.2%. In a bivariate analysis a significant correlation was found between cardiac death and work capacity (p less than 0.02), vessel involvement (greater than 50% stenosis), left ventricular impairment as evaluated by ventriculography (p less than 0.001) and heart volume enlargement (p less than 0.01). There was no significant correlation between the degree of ST-segment depression and ventricular premature depolarisation (during exercise test). Complex results from non-invasive approaches can identify groups with good and bad prognosis as effectively as invasive techniques.