OBJECTIVE Diagnostic methods validation and incidence estimation of silent myocardial ischemia in patients with previous myocardial infarction under 45 years. METHODS Prospective in comparison with a healthy group. METHODS Cardiac outpatients follow-up at Faro's District Hospital. METHODS A random group of 23 outpatients (GI) under 45 years, with previous myocardial infarction. A second group of healthy volunteers (GII) similar in age and sex. METHODS After discontinuing therapy, a maximal treadmill exercise test (E.T.) was performed in both groups, using the Bruce protocol. A ST segment depression greater than or equal to 1 mm, measured 80 ms after J point was the positive criteria. Simultaneously a 24 h Holter recording was obtained using a two channel real time recorder. ST segment depression greater than or equal to 1 mm, measured 80 ms after J point and lasting over 60 s., was the positive criteria. Patients with left bundle branch block or left ventricular hypertrophy criteria were excluded. Concerning ventricular arrhythmias only repetitive forms were considered. RESULTS Ten E.T. (43.5%) resulted positive in GI. Simultaneous Holter recording was positive in nine patients (one false negative). From the 13 patients with negative E.T., 12 had negative Holter recordings (one false positive). Every E.T. and simultaneous Holter resulted negative in GII. Silent ischemia was detected in eight GI patients (34.8%) all of them belonging to the subgroup with positive E.T. In four patients the silent ischemia was detected by asymptomatic E.T., and simultaneous Holter. The remaining four patients had silent ischemia diagnosed on the subsequent Holter. Silent ischemia episodes were not detected in the subgroup of 13 patients with negative E.T. Between those two subgroups it is highly significant (p less than 0.001) the difference in the incidence of silent ischemia. The patients with silent ischemia recorded an average of 6.5 episodes/patient/day mainly in day time (p less than 0.001). Episodes of silent ischemia were more frequent in the subgroup of patients with asymptomatic positive E.T. than in the subgroup of positive E.T. with pain (p less than 0.004). It was not confirmed any significant difference in the incidence of ventricular arrhythmias among patients with or without silent ischemia. CONCLUSIONS We have verified an high incidence of silent ischemia in a group of patients with previous myocardial infarction. Holter's electrocardiographic monitoring has a high concordance with E.T. results, when performed simultaneously. In fact, it does not provide any significant additional information since every patient with silent ischemia had positive E.T., but can be complementary in the evaluation of the total ischemic burden. It has been checked that episodes of silent ischemia have a major incidence at day time, appearing more often in patients with asymptomatic positive E.T.