The present study aimed to determine whether amino-terminal pro-brain natriuretic peptide (NT-pro-BNP) predicts intensive care unit (ICU) mortality in a cohort of general, noncardiac, critically ill patients. To this end, a total of 233 consecutive mechanically ventilated patients (109 men) having a median age of 60 years and a wide range in admitting diagnoses, including medical (n = 118), surgical (n = 83), and multiple trauma (n = 32) cases were prospectively studied. Median Acute Physiology and Chronic Health Evaluation (APACHE) II and Sequential Organ Failure Assessment scores on ICU admission were 16 and 9, respectively. The study end point was ICU outcome. Blood samples were drawn on admission in the ICU and on postadmission days 1 and 2 to determine NT-pro-BNP levels. In a subgroup (n = 77), admission proinflammatory and anti-inflammatory cytokine levels, including TNF-alpha, IL-6, and IL-10, were also measured. Nonsurvivors (n = 98) had significantly higher NT-pro-BNP levels than survivors (n = 135) on admission in the ICU (2,074 vs. 283 pg/mL; P < 0.001), on day 1 (2,197 vs. 221 pg/mL; P < 0.001), and on day 2 (2,726 vs. 139 pg/mL; P < 0.001). Median values for TNF-alpha, IL-6, and IL-10 were 3.70, 131.57, and 111.88 pg/mL, respectively. Receiver operating characteristic analysis showed that the area under the receiver operating characteristic curve in predicting ICU mortality was 0.70 for APACHE II and 0.77 for admission NT-pro-BNP (P = 0.08). The cutoff in admission NT-pro-BNP that best predicted outcome was 941 pg/mL. Multiple logistic regression analysis revealed that APACHE II score (odds ratio, 1.06; P = 0.007) and the best cutoff point in admission NT-pro-BNP (odds ratio, 7.74; P < 0.001) independently predicted ICU mortality, even if cytokines were entered in the analysis. In conclusion, plasma NT-pro-BNP is frequently raised in noncardiac, mixed, critically ill patients, and nonsurvivors have consistently higher levels than survivors. Elevated admission NT-pro-BNP represents an independent predictor for poor ICU outcome in the presence of clinical severity scores.