Background A deleterious effect of complete left bundle-branch block (LBBB) on left ventricular function has been established. Nevertheless, the independent effect of a widened QRS on mortality rate in congestive heart failure (CHF) is still controversial. Therefore, we carried out this analysis to determine whether LBBB is an independent predictor of mortality in CHF. Methods and Results We analyzed the large Italian Network on CHF Registry of unselected outpatients with CHF of different causes. The registry was established by the Italian Association of Hospital Cardiologists in 1995. Complete 1-year follow-up data were available for 5517 patients. The main underlying cardiac diagnosis was ischemic heart disease in 2512 patients (45.6%), dilated cardiomyopathy in 1988 patients (36.0%), and hypertensive heart disease in 714 patients (12.9%). Other causes were recorded in the remaining 303 cases (5.5%). LBBB was present in 1391 patients (25.2%) and was associated with an increased 1-year mortality rate from any cause (hazard ratio, 1.70; 95% confidence interval, 1.41 to 2.05) and sudden death (hazard ratio, 1.58; 95% confidence interval, 1.21 to 2.06). Multivariate analysis showed that such an increased risk was still significant after adjusting for age, underlying cardiac disease, indicators of CHF severity, and prescription of angiotensin-converting enzyme inhibitors and β-blockers. Conclusion LBBB is an unfavorable prognostic marker in patients with CHF. The negative effect is independent of age, CHF severity, and drug prescriptions. These data may support the rationale of randomized trials to verify the effects on mortality rate of ventricular resynchronization with multisite cardiac pacing in patients with CHF and LBBB. (Am Heart J 2002;143:398-405.)