The alveolar nitrogen slope (PIII), closing volume (CV), and closing capacity (CC) were measured by the single-breath nitrogen washout method (SBN2) in a group of 187 healthy children and adolescents (92 boys, 95 girls), 10 to 16 yr old, from the general population of Lorraine, France. The test was performed using a computerized system, which also made the calculations. About one out of five healthy subjects in this population were unable to satisfactorily perform the test; the failure rate was the same for the two sexes (20% in boys, 21.5% in girls) and significantly higher in younger children (26.6 and 14.5% for children under and over the age of 13, respectively; p = 0.03). The distribution of results was skewed for PIII and practically normal for log PIII, CV, VC, and CV/VC or CC/TLC ratios. PIII was highly significantly, inversely related to anthropometric variables; the highest coefficient was that for the age-weight interaction term in boys (= r -0.57 for PIII, -0.62 for log PIII) and for weight in girls (r = -0.57 for both PIII and log PIII). Because the anthropometric variables were strongly interrelated (r between 0.45 and 0.79), multiple regressions did not materially improve the prediction of PIII. In simple regression, weight alone explained 36% of the variability of log PIII in boys and 32% in girls. The mean PIII was significantly higher in girls as compared to boys (1.14 +/- 0.38 versus 0.98 +/- 0.17% N2/L, p = 0.02); CV and CC in milliliters were related to body build as other lung volumes; the CV/VC in girls and CC/TLC ratio in both sexes were not related to anthropometric variables. In boys, CV/VC decreased significantly with height (p = 0.035 for CV/VC versus height3).