Renal function and urea are frequently abnormal in patients with heart failure (HF) and are predictive of increased mortality. The relative importance of each parameter is less clear. We prospectively compared the predictive value of renal function and serum urea on clinical outcome in patients with HF. Patients hospitalized with definite clinical diagnosis of HF (n = 355) were followed for short-term (1 yr) and long-term (mean, 6.5 yr) survival and HF rehospitalization. Increasing tertiles of discharge estimated glomerular filtration rate (eGFR) were an independent predictor of increased long-term survival (hazard ratio [HR], 0.65; 95% confidence interval [CI], 0.47-0.91; p = 0.01) but not short-term survival. Admission and discharge serum urea and blood urea nitrogen (BUN)/creatinine ratio were predictors of reduced short- and long-term survival on multivariate Cox regression analysis. Increasing tertiles of discharge urea were a predictor of reduced 1-year survival (HR, 2.13; 95% CI, 1.21-3.73; p = 0.009) and long-term survival (HR, 1.93; 95% CI, 1.37-2.71; p < 0.0001). Multivariate analysis including discharge eGFR and serum urea demonstrated that only serum urea remained a significant predictor of long-term survival; however, eGFR and BUN/creatinine ratio were both independently predictive of survival. Urea was more discriminative than eGFR in predicting long-term survival by area under the receiver operating characteristic curve (0.803 vs. 0.787; p = 0.01). Increasing tertiles of discharge serum urea and BUN/creatinine were independent predictors of HF rehospitalization and combined death and HF rehospitalization. This study suggests that serum urea is a more powerful predictor of survival than eGFR in patients with HF. This may be due to urea's relation to key biological parameters including renal, hemodynamic, and neurohormonal parameters pertaining to the overall clinical status of the patient with chronic HF.