We prospectively evaluated whether the combination of admission measurements of a marker for myocardial cell injury and a marker for left ventricular overload would effectively risk stratify patients with acutely decompensated heart failure. We measured serum concentrations of cardiac troponin T (cTnT) using a second-generation assay, as well as serum cardiac troponin I (cTnI) and plasma atrial and brain natriuretic peptide (BNP) concentrations on admission in 98 consecutive patients hospitalized for worsening chronic heart failure (mean age 69 years; 5 patients were in New York Heart Association functional class II, 35 were in class III, and 58 patients were in class IV). During a mean follow-up period of 451 days, there were 37 cardiac events, including 21 cardiac deaths (14 in-hospital deaths) and 16 readmissions for worsening heart failure. In a stepwise Cox regression analysis, including these biochemical markers, age, sex, functional class, and left ventricular ejection fraction, cTnT, and BNP were found to be significantly independent predictors of both cardiac death (p <0.05) and cardiac events (p <0.01). A cTnT >0.033 microg/L and/or a BNP >440 pg/ml on admission was correlated with an incremental increase in in-hospital cardiac mortality, overall cardiac mortality, and cardiac event rate. Kaplan-Meier analysis revealed that this combination could reliably stratify the patients into low-, intermediate-, and high-risk groups for cardiac events. Measuring the combination of admission concentrations of cTnT and BNP may be a highly effective means of risk stratification of patients hospitalized for worsening chronic heart failure.