This study was conducted to assess the ability of preoperative B-type natriuretic peptide levels to predict medium-term mortality in patients who undergo major noncardiac surgery. During a median of 654 days of follow-up, 33 patients from a total cohort of 204 patients (16%) died, 17 from cardiovascular causes. The optimal cutoff in this cohort, determined using a receiver-operating characteristic curve, was >35 pg . ml(-1). This was associated with a 3.5-fold increase in the hazard of death (p = 0.001) and a 6.9-fold increase in the hazard of cardiovascular mortality (p = 0.003). In conclusion, these findings extend recent work demonstrating that B-type natriuretic peptide levels obtained before major noncardiac surgery can be used to predict perioperative morbidity and indicate that they also forecast medium-term mortality, particularly cardiovascular death.