Studies of electrocardiographic predictors of mortality in patients with chronic heart failure have reached varying conclusions. Differences in the characteristics of the patients studied may explain the conflicting results regarding both a prolonged QRS and an abnormal signal-averaged electrocardiogram (SAE). We therefore investigated the impact of the etiology of heart failure on the prognostic importance of a prolonged QRS and an abnormal SAE in 200 patients with heart failure. Patients were categorized according to etiology of heart failure and electrocardiographic parameters. The mortality of patients with a prolonged QRS was compared with mortality in those with both abnormal and normal SAEs. This was done for the entire group, and separately for those with ischemic and those with nonischemic cardiomyopathy. The mean follow-up was 18.8 months. Nonischemic patients with a prolonged QRS had significantly worse survival than other patients. However, nonischemic patients with an abnormal SAE did not have a worse prognosis than patients with a normal SAE. One-year survival of patients with a prolonged QRS was 71%, compared with 98% in patients with a normal and 87% in patients with an abnormal SAE (p < 0.05). In contrast, a prolonged QRS was not a predictor of poor prognosis in patients with ischemic cardiomyopathy (81% one year mortality). Patients with ischemic cardiomyopathy and an abnormal SAE tended to have a poorer survival than patients with a normal SAE (73% and 81% one year mortality, respectively). Thus, the etiology of heart failure affects the prognostic importance of both a prolonged QRS and an abnormal SAE.