Rapid-assay biomarkers may predict outcomes in patients with decompensated heart failure (HF). This study assessed whether rapid-assay B-type natriuretic peptide (BNP) and troponin I predicts length of stay and mortality and correlates with pulmonary artery catheter (PAC)-derived hemodynamics in patients hospitalized with acute HF. There were 141 nonconsecutive patients in this prospective cohort study of the Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheterization Effectiveness (ESCAPE), a randomized trial testing PACs in 433 patients with severe decompensated HF. Biomarkers were drawn at baseline and discharge and when the first, second, and final hemodynamics were obtained in 69 patients randomly assigned to PACs. Cox analysis was used to model mortality, length of stay, and rehospitalization, and Pearson's correlations were used to describe the relation among BNP, troponin I, and PAC-derived hemodynamics. The median (25th percentile, 75th percentile) BNP levels were 783 pg/ml (329, 1,565) at baseline and 468 pg/ml (240, 946) at discharge. After treatment for HF, the median BNP level decreased by 144 pg/ml (-653, 55; p = 0.004). Patients with baseline BNP levels >1,500 pg/ml had greater mortality at 6 months and almost twice the length of stay as patients with BNP levels <500 pg/ml (10.1 vs 5.7 days, p = 0.002). Troponin I did not predict these outcomes. First BNP correlated modestly with first right atrial pressure (r = 0.47, p = 0.005) and first pulmonary capillary wedge pressure (r = 0.54, p = 0.001). Final BNP correlated modestly with final right atrial pressure (r = 0.63, p = 0.001). In conclusion, patients with BNP >1,500 pg/ml had greater mortality and longer length of stay than patients with BNP <500 pg/ml. BNP decreased after hospitalization, but correlated modestly with PAC-derived hemodynamics. Rapid-assay BNP may provide information that helps physicians decide when to pursue more aggressive and invasive therapies.