The risk of cardiac arrest is increased during strenuous physical exercise in patients with stable coronary artery disease (CAD). Because premonitoring symptoms are rarely observed, silent myocardial ischemia may represent the pathophysiological basis for the induction of malignant ventricular arrhythmias. Holter monitoring was, therefore, performed in 40 consecutive patients entering a randomized intervention trial on progression of CAD. In 20 of 21 participants (95%) in the intervention program greater than or equal to 1 episode of silent myocardial ischemia was observed during the initial training session. The mean duration of silent myocardial ischemia per patient was 25 +/- 13 min/hr of training session. During normal daily activity only 5 patients (24%) experienced greater than or equal to 1 episode of silent myocardial ischemia (p less than 0.001) yielding a mean duration of 0.6 +/- 1.3 minutes of silent myocardial ischemia/hr of ordinary activity per patient (p less than 0.001 vs training session). During a control period of 24 hours without exercise training the incidence (33%) and mean duration of silent myocardial ischemia (0.8 +/- 2.1 min/hr/patient) were similar to those during normal daily activity on the day of the training session. During the training session the occurrence of frequent or repetitive ventricular arrhythmias was related to 10 silent myocardial ischemia episodes detected in 5 patients. During normal daily activity in 1 patient only was the onset of malignant ventricular arrhythmias associated with silent myocardial ischemia (p less than 0.05). Conditions and results of the Holter studies in the control group patients were comparable to those of the patients in the intervention group on the day without physical exercise.(ABSTRACT TRUNCATED AT 250 WORDS)