Objective. Air diplacement plethysmography (ADP) has become increasingly popular to assess body composition in children. The aim of this study was to compare ADP with deuterium dilution and to investigate the effect of using child-specific prediction equations to correct raw body volume from ADP for thoracic gas volume (TGV) and body surface area (BSA). Methods. Thirty-seven healthy Dutch children (17 girls, 20 boys) aged 10-13 years were recruited. Body volume was measured using the Bod Pod(R). Both adult and child-specific prediction equations were used to correct raw body volume from the Bod Pod(R) for TGV and BSA. Total body water (TBW) was assessed using deuterium dilution. Child-specific densities and hydration fractions of fat-free mass were used to convert body volume and TBW to percentage fat mass (%FM). Correlation and agreement between methods was assessed using linear regression analysis and Bland-Altman plots. Results. Despite a high correlation between the Bod Pod(R) and deuterium dilution (R=0.91, p<0.001), a significant difference was found between %FM obtained using the Bod Pod(R) and deuterium dilution (p<0.001), regardless of the equation used to correct raw body volume for TGV and BSA. Bland-Altman plots showed a systematic bias towards a smaller difference between techniques at higher %FM. Conclusion. Significant differences in %FM were found between both methods. Given the underlying assumptions to translate body volume or TBW to %FM, it is recommended to use a 3- or 4-compartment model when assessing body composition in children.