Forced oscillatory respiratory resistance was measured in 138 healthy children aged 2- to 16-years-old using a commercial unit. Regression analysis was performed using the resistance measurements obtained at frequencies of 6 Hz and 26 Hz and averages of measurements obtained at frequencies of from 6-26 Hz, 6-10 Hz, and 22-26 Hz. The analysis suggested that regression curves and 95% confidence intervals obtained using a quadratic model with height as the independent variable provided the best prediction of normal values and their ranges. Three independent sets of resistance measurements, which were obtained for all 138 subjects, were used to calculate the average individual coefficients of variation (CV) for the five resistance parameters noted above; these ranged from 9 to 13%. This analysis suggested that the expected CV of repeated measurements would be about 10%. Finally, values from 13 children who had asthma with mild bronchial obstruction (FEV1 less than 80% of predicted) were compared with the normal values derived from the regression curves. Values from all but two of the children fell within the reference intervals, suggesting that FEV1 may be a more sensitive measure of obstruction than forced oscillatory resistance. The main applications of this new approach may be in the evaluation of young children, in following individual patients, and in bronchodilation and bronchial challenge tests.