The aim of the present study was the investigation of N-terminal pro-brain natriuretic peptide (NT-proBNP) in the pediatric population. This is essential for adequate monitoring and classification of pediatric patients with heart disease, but no consistent data are available yet. In addition, the comparability of two commercially available NT-proBNP assays and the inter-laboratory variability for the most suitable one were assessed. For this purpose, 408 subjects (1-29 years) were included. NT-proBNP was determined with a non-competitive electrochemiluminescent immunoassay (Roche NT-proBNP; n = 402) and a competitive enzyme-immunoassay (Biomedica NT-proBNP; n = 402). Inter-laboratory variability was evaluated for the Roche assay by stepwise inclusion of four and 11 centers throughout Germany, respectively. Roche NT-proBNP ranged from 5.0 to 391.5 ng/L, with higher values for younger children. The 97.5th (75th) percentile curve ranged from 319.9 ng/L (231.2 ng/L, 1-3 years) to 114.9 ng/L (53.3 ng/L, 18 years). In contrast, Biomedica NT-proBNP ranged from 253.7 to 7602.8 ng/L, with no significant age dependency. The mean difference between the assays was 1649.7 ng/L (95% confidence interval 1546.3-1753.1 ng/L). Inter-laboratory variability ranged from 6.5% to 3.8%, covering a range from 51.3 to 6618.1 ng/L. The assay seems to influence the interpretation of resulting NT-proBNP values and therefore has to be chosen carefully. For the monitoring and classification of pediatric patients with congenital heart disease, age-based NT-proBNP values should be used.