The Cardiac Arrhythmia Pilot Study (CAPS) was a randomized, double-blind trial of antiarrhythmic drugs (encainide, flecainide, moricizine, imipramine and placebo) in 502 patients with at least 10 ventricular premature complexes/hour, 6 to 60 days after acute myocardial infarction. CAPS tested the feasibility of performing a larger study to determine if suppression of ventricular ectopic activity after acute myocardial infarction could improve survival. Patients in CAPS were followed for 1 year. All death or cardiac arrest events were evaluated by at least 2 investigators using a classification scheme that characterized the underlying mechanism as cardiac arrhythmic, cardiac nonarrhythmic or noncardiac. Forty-five patients (9%) died or had cardiac arrest during the 1-year follow-up, 29 (64%) within 1 hour from the onset of symptoms and 16 greater than 1 hour from the onset of symptoms. Twenty-three deaths (51%) were classified as arrhythmic, 19 (42%) as nonarrhythmic and 3 (7%) as noncardiac. Acute myocardial ischemia or infarction was associated with the death/cardiac arrest event in 16 patients (36%), 8 in the arrhythmic death group. Discrepancies in classification among reviewers were particularly common in patients with long-standing symptoms of congestive heart failure, in whom it was frequently difficult to identify the precise moment of the onset of symptoms in the death/cardiac arrest event. Using only the temporal relation of symptoms to categorize deaths or cardiac arrests, the mechanism of 12 (27%) of the 45 patients was in disagreement with the classification based on the Events Committee review. Classification of death as sudden or nonsudden is not equivalent to the classification of death as arrhythmic or nonarrhythmic.