Heart failure is a disease characterized by high prevalence and mortality, and frequent rehospitalizations. The aim of this study is to investigate the prognostic power of combining brain natriuretic peptide (BNP) and congestion status detected by bioelectrical impedance vector analysis (BIVA) in acute heart failure patients. This is an observational, prospective, and a multicentre study. BNP assessment was measured upon hospital arrival, while BIVA analysis was obtained at the time of discharge. Cardiovascular deaths were evaluated at 90 days by a follow up phone call. 292 patients were enrolled. Compared to survivors, BNP was higher in the non-survivors group (mean value 838 vs 515 pg/ml, p < 0.001). At discharge, BIVA shows a statistically significant difference in hydration status between survivors and non-survivors [respectively, hydration index (HI) 85 vs 74, p < 0.001; reactance (Xc) 26.7 vs 37, p < 0.001; resistance (R) 445 vs 503, p < 0.01)]. Discharge BIVA shows a prognostic value in predicting cardiovascular death [HI: area under the curve (AUC) 0.715, 95% confidence interval (95% CI) 0.65-0.76; p < 0.004; Xc: AUC 0.712, 95% CI 0.655-0.76, p < 0.007; R: AUC 0.65, 95% CI 0.29-0.706, p < 0.0247]. The combination of BIVA with BNP gives a greater prognostic power for cardiovascular mortality [combined receiving operating characteristic (ROC): AUC 0.74; 95% CI 0.68-0.79; p < 0.001]. In acute heart failure patients, higher BNP levels upon hospital admission, and congestion detected by BIVA at discharge have a significant predictive value for 90 days cardiovascular mortality. The combined use of admission BNP and BIVA discharge seems to be a useful tool for increasing prognostic power in these patients.