Neurohormonal activation is prevalent in adults with congenital heart disease, but its relation to outcome remains unknown. B-type natriuretic peptide (BNP) and atrial natriuretic peptide (ANP) were measured prospectively in 49 patients with adult congenital heart disease, who were followed up for a median of 7.9 years (interquartile range 7.7 to 8.2). Cox proportional hazards regression analysis was used to determine the relation of BNP and ANP concentrations to all-cause mortality. The mean age at baseline was 33.9 +/- 11.3 years, and 46.9% of patients were men. Most patients (77.5%) were symptomatic (20.4% had New York Heart Association class III), 10 (20.4%) were cyanotic, and 28 (57.1%) had systemic ventricular dysfunction (moderate or severe in 18.4%). The median concentration of BNP was 52.7 pg/ml (interquartile range 39.1 to 115.4) and of ANP was 47.4 pg/ml (interquartile range 19.7 to 112.8). Of the 49 patients, 11 (22.4%) died during the follow-up period. Both BNP and ANP were strong predictors of mortality (hazard ratio per 100-pg/ml increase 1.80, 95% confidence interval 1.38 to 2.34, p <0.0001; and hazard ratio per 100-pg/ml increase 1.21, 95% confidence interval 1.12 to 1.32, p <0.0001, respectively). A BNP value >78 pg/ml predicted death with a sensitivity of 100% and specificity of 76.3% (area under the curve 0.91, p = 0.0001). An ANP value of >146 pg/ml predicted death with a sensitivity of 72.7% and specificity 94.7% (area under the curve 0.89, p = 0.0001). No patients with a BNP level <78 pg/ml died during the follow-up period. In conclusion, the BNP and ANP levels strongly predicted death in symptomatic ambulatory patients with adult congenital heart disease during mid-term follow-up and could be used as a simple clinical marker for risk stratification in this population.