The aim of this study was to compare tidal volume (VT) derived from the Qualitative Diagnostic Calibration (QDC) method (VT,QDC) with measurements from pneumotachography (VT,PN,T) in anaesthetized infants. Measurements were made during spontaneous (SV) and intermittent positive pressure (IPPV) ventilation, sighs and airway occlusions. The VT,DIF was the difference between VT,QDC and VT,PNT (%VT). The contribution of the ribcage (rc) to VT,QDC (%rc) and the thoracoabdominal phase lag were also derived. Twenty-eight infants, mean (SD) age 14.0 (6.2) months were studied. VT,QDC represented VT,PNT most closely when > or = 20 breaths were analysed. There was close agreement during SV immediately after the calibration period (95% limits of agreement (LA; QDC - PNT) -23, 3.0%). The 95% LA increased to -9.6, 10.2% after 10 min. Accuracy diminished during IPPV (95% LA -38, 31%), and sighs. During airway occlusions, when VT,PNT was zero, the 95% LA were -63, 4.1 mL x kg(-1). Mean phase lag was 36 and 2%, respectively, during SV and IPPV (p<0.05). The %rc appeared to be overestimated, being in excess of 50% in infants under 12 months. The Qualitative Diagnostic Calibration method used to estimate tidal volume in anaesthetized infants was limited by the need to analyse > or = 20 breaths and by a loss of within-subject accuracy if measurement conditions or pattern of breathing changed.