Cardiac arrest during swimming accounts for a considerable number of deaths during physical exercise in patients with coronary artery disease. A link between ST-segment depression and cardiac arrest has been observed in previous studies. In this study, exercise-induced myocardial ischemia was assessed in 23 patients with coronary artery disease by bipolar Holter monitoring during swimming, jogging, and treadmill testing. During treadmill testing, Holter monitoring and standard electrocardiograms were simultaneously recorded. Detection of ST-segment depression during swimming was standardized in a group of normal volunteers (n = 7). All patients with silent myocardial ischemia (n = 8) documented by thallium-201 scintigraphy had ST-segment depression during treadmill testing and swimming when recorded by Holter monitoring, whereas the standard electrocardiogram during treadmill testing was negative in 5 patients. Heart rate at 1 mm ST-segment depression was significantly lower during swimming (110 +/- 11 beats/min) than during treadmill testing (documented by standard electrocardiogram) (133 +/- 23 beats/min, p < 0.002) and jogging (125 +/- 21 beats/min, p < 0.03). However, there was no significant difference in heart rate at onset of angina pectoris in symptomatic patients, suggesting a delayed sensation of ischemic symptoms during swimming. The only clinical event in our group during 8 years of swimming occurred during this study. One patient with silent myocardial ischemia developed ST-segment depression during swimming that degenerated into ventricular fibrillation, requiring resuscitation. Therefore, Holter monitoring can be considered a valuable addition in identifying patients with silent myocardial ischemia during swimming, and thus identifying patients at risk for exertion-related life-threatening ventricular tachyarrhythmias.