Sudden arrhythmic cardiac death is a major unresolved health problem, yet there is no agreement on the chronologic definition of sudden death. This retrospective study investigates the frequency distribution of the chronology of the terminal cardiac event in a large postinfarction population and identifies factors associated with instantaneous (< 1 minute) cardiac death. This study involved 229 patients enrolled in the Multicenter Diltiazem Post-infarction Trial who died during 2-year follow-up and had quantitative information on the chronology of the terminal event. Thirty-two percent of the cardiac deaths occurred instantaneously. Patients who died instantaneously were more likely (p < 0.05) to be men, to have a baseline ejection fraction < 0.40, and to have frequent (> or = 10/hour) and repetitive (> or = 3 in a row) ventricular ectopic complexes (VECs) on an ambulatory electrocardiogram than those who did not die instantaneously. Patients who died instantaneously received more digitalis and class IA antiarrhythmic agents and less beta blockers in the week before death than those dying noninstantaneously. Logistic regression analysis identified 3 independent factors that differentiated instantaneous from noninstantaneous death (relative risk; 95% confidence interval): frequent VECs (2.15; 1.11 to 4.17); digitalis (2.57; 1.31 to 5.06); and no beta blocker medication (2.90; 1.09 to 7.75). Instantaneous death (within 1 minute) was responsible for almost one third of the cardiac deaths that occurred in this postinfarction population. Frequent VECs, digitalis, and absence of beta-blocker therapy distinguished patients who died instantaneously from those who died noninstantaneously.