The gold standard for assessing quality of forced expiratory manoeuvres is visual inspection by an expert. American Thoracic Society/European Respiratory Society numerical quality criteria (NQC) include back-extrapolated volume (BEV), repeatability and forced expiratory time (FET). Equipment currently available provides feedback tempting the investigator to use NQC as pass-fail criterion. To investigate whether using NQC instead of visual acceptability is a valid option, we analysed data from a multicentre national reference study in Germany of children aged 4-18 years. Spirometry was performed under field conditions. Receiver operating characteristic analysis was used to assess performance of BEV, repeatability, FET and a combination thereof in relation to visual acceptability. We included data from 3133 healthy Caucasians in the analyses; 72% delivered at least two visually acceptable manoeuvres. Of these, 59% would have been rejected based on combined NQC, mainly because the FET criterion was not feasible. Specificity of the NQC was generally low (BEV 10%, repeatability 30% and FET 50%). Receiver operating characteristic analysis showed that a combination of the three measures could reach at best a sensitivity of 90% and specificity of 56%. We conclude that visual control is mandatory and NQC may help obtain the best possible results, but a fixed cut-off for FET should be abandoned.