The prognostic significance of ambulatory ischemia, alone and in relation to ischemia during exercise was assessed in 686 patients (475 men) with chronic stable angina pectoris taking part in the Angina Prognosis Study In Stockholm (APSIS), who had 24-hour ambulatory electrocardiographic registrations and exercise tests at baseline (n = 678) and after 1 month (n = 607) of double-blind treatment with metoprolol or verapamil. Ambulatory electrocardiograms were analyzed for ventricular premature complexes and ST-segment depression. During a median follow-up of 40 months, 29 patients died of cardiovascular (CV) causes, 27 had a nonfatal myocardial infarction, and 89 underwent revascularization. Patients with CV death had more episodes (median 5 vs. 1; p<0.01) and longer median duration (24 vs. 3 minutes; p<0.01) of ST-segment depression than patients without events. For those who had undergone revascularization, the duration was also longer (12 vs. 3 minutes; p<0.05). In a multivariate Cox model including sex, history of previous myocardial infarction, hypertension, and diabetes, the duration of ST-segment depression independently predicted CV death. When exercise testing was included, ambulatory ischemia carried additional prognostic information only among patients with ST-segment depression > or =2 mm during exercise. When the treatment given and treatment effects on ambulatory ischemia were added to the Cox model, no significant impact on prognosis was found. Ventricular premature complexes carried no prognostic information. Thus, in patients with stable angina pectoris, ischemia during ambulatory monitoring showed independent prognostic importance regarding CV death. Ambulatory electrocardiographic monitoring and exercise testing provide complementary information, but only among patients with marked ischemia during exercise. Treatment reduced ambulatory ischemia, but the short-term treatment effects did not significantly influence prognosis.