Background: Acute Heart failure (AHF) is the one of the most common cardiovascular disorders with high morbidity and mortality. Numerous parameters have been evaluated as predictors for AHF. The purpose of this study was to determine the prognostic factors of long-term clinical outcomes in patients with AHF using C5.0 decision-tree algorithm. Methods: We analyzed 355 consecutive patients (149 males, age 68.5짹12.5 years) with AHF. Cardiovascular events defined as readmission related heart failure (HF) or cardiovascular death within 2 years. C5.0 decision-tree algorithm was used as a tool to identify predictors for long-term clinical outcomes in patients with AHF. Univariate and multivariate Cox proportional analysis was performed to determine the prognostic value of predictors identified by C5.0 decision-tree algorithm. Results: Of the 355 patients, 41 (11.5%) died and 104 (29.3%) were re-admitted for HF during the 2-year follow-up period. In C5.0 decision tree-algorithm and univariate Cox analysis, presence of diabetes and previous HF, blood urea nitrogen (BUN) >26 mg/dl, serum creatinine >1.3 mg/dl and N-terminal Pro B-type natriuretic peptide (NT-ProBNP) >8128.3 pg/ml were associated with combined cardiovascular events. Multivariate analysis showed only BUN>26 mg/dl was independently associated with poor long-term outcomes. The patients with NT-ProBNP>8128.3 pg/ml showed tendency of higher incidence of long-term cardiovascular events. Conclusion: Data analysis using C5.0 decision- tree algorithm showed that increased BUN (>26 mg/dl) appears to be an independent poor prognostic factor for long-term cardiovascular events in patients with AHF.