This retrospective study examines the prognostic significance of exercise-induced ventricular arrhythmia in patients with stable coronary artery disease (CAD) who were included in the multicenter patient registry of the Coronary Artery Surgery Study. The population is composed of 1,486 patients selected from 1975 to 1979 and followed an average of 4.3 years. All underwent a standard Bruce exercise test and had CAD by cardiac catheterization at entry. Patients were classified into group I or II depending on whether they had minimal or significant CAD. (Significant CAD was defined as 70% or greater diameter reduction in any major coronary artery or 50% or greater narrowing in the left main artery.) They were further subclassified into groups A or B depending on whether or not they had exercise-induced ventricular arrhythmia. Groups IA (16 patients) and IB (229 patients) had similar clinical and angiographic characteristics except for the average ejection fraction (EF), which was 50% for group IA and 64% for group IB (p less than 0.05). Group IIA (130 patients) had a higher prevalence of previous myocardial infarction, a lower mean EF and a higher proportion of patients with at least 2 coronary arteries significantly narrowed than group IIB (1,111 patients). The 5-year event-free survival was not influenced by the presence of exercise-induced ventricular arrhythmia; it was 76 and 88% in groups IA and IB, respectively (difference not significant), and 71 and 76% in groups IIA and IIB, respectively (difference not significant).(ABSTRACT TRUNCATED AT 250 WORDS)