Physiologic studies have suggested that anemia could adversely affect the cardiovascular condition of patients with heart failure. Yet, the prognostic importance of this treatable condition is not well established by epidemiologic studies. We sought to determine the prognostic value of hematocrit level in a cohort of elderly patients hospitalized with heart failure. We studied a consecutive sample of 2281 patients aged 65 years or older who had been admitted with a principal discharge diagnosis of heart failure. Multivariate Cox proportional hazards regression was conducted to test whether hematocrit level was an independent predictor of 1-year mortality and of hospital readmission. The mean (+/- SD) age of the patients was 79 +/- 8 years; 58% (n = 1324) were women. Their median hematocrit was 38% (25th to 75th percentile, 33% to 42%). Lower hematocrits were associated with a higher mortality. After adjusting for demographic and clinical factors, each 1% lower hematocrit was associated with a 2% greater 1-year mortality (P = 0.007). Compared with patients with a hematocrit >42%, those with a hematocrit < or =27% had a 40% greater 1-year mortality (hazard ratio [HR] = 1.40; 95% confidence interval [CI]: 1.02 to 1.92; P = 0.04). This increased risk was similar to that conferred by traditional risk factors, including a left ventricular ejection fraction < or =20% (HR = 1.50; 95% CI: 1.20 to 1.86). Lower hematocrits were also associated with a greater risk of hospital readmission. Anemia is associated with an increased risk of death and rehospitalization in older patients with heart failure. Whether anemia is a direct cause of worse outcomes, or a marker for other causal factors, is not known.