Many ventilators measure expired tidal volume (VT) without compensation either for the compliance of the ventilator circuit or for variations in the circuit setup. We hypothesized that the exhaled VT measured with a conventional ventilator at the expiratory valve would differ significantly from the exhaled VT measured with a pneumotachometer placed at the endotracheal tube. To investigate this we studied 98 infants and children requiring conventional ventilation. We used linear regression analysis to compare the VT obtained with the pneumotachometer with the ventilator-measured volume. An additional comparison was made between the pneumotachometer volume and a calculated effective VT. For infant circuits (n = 70), our analysis revealed a poor correlation between the expiratory VT measured with the pneumotachometer and the ventilator-measured volume (r(2) = 0.54). Similarly, the expiratory VT measured with the pneumotachometer did not correlate with the calculated effective volume (r(2) = 0.58). For pediatric circuits (n = 28), there was improved correlation between the expiratory VT measured with the pneumotachometer and both the ventilator-measured volume and the calculated effective VT (r(2) = 0.84 and r(2) = 0.85, respectively). The data demonstrate a significant discrepancy between expiratory VT measured at a ventilator and that measured with a pneumotachometer placed at the endotracheal tube in infants. Correcting for the compliance of the ventilator circuit by calculating the effective VT did not alter this discrepancy. In conventionally ventilated infants, exhaled VT should be determined with a pneumotachometer placed at the airway.