Earlier studies have suggested that patients exhibiting late (> 24 h) ventricular fibrillation during acute myocardial infarction had a poorer outcome in comparison to myocardial infarction patients with early (< 24 h) ventricular fibrillation. Between August 1981 and July 1983, 5839 consecutive patients with acute myocardial infarction were hospitalized in 13 out of 21 operating coronary care units in Israel. Demographic and medical data were collected from hospitalization charts and during 1 year of follow-up. Mortality assessment was done for 99% of hospital survivors up to mid-1988 (mean, 5.5 years). The incidence of ventricular fibrillation in the SPRINT Registry was 6% (371/5839). Patients with ventricular fibrillation in the setting of cardiogenic shock (n = 107) were excluded from analysis. Patients with late ventricular fibrillation (n = 109; 41%) were older and had a more complicated hospital course than patients with early ventricular fibrillation (n = 155; 59%). In-hospital and 1-year post-discharge mortality were significantly higher in patients with late ventricular fibrillation (63% and 17%) as compared to patients with early ventricular fibrillation (26% and 4%, respectively; P < 0.05 for each). This difference vanished 5 years after hospital discharge. After multiple logistic regression analysis late occurrence of ventricular fibrillation emerged as an independent predictor of increased in-hospital mortality (Odds ratio, 4.29; 95% confidence interval, 2.11-8.74) but not for subsequent death. Patients with late ventricular fibrillation during the hospital course of acute myocardial infarction had a poorer immediate and subsequent outcome in comparison to patients with early ventricular fibrillation.