Air leak around a tracheal tube (TT) during mechanical ventilation is likely to occur during the inspiratory phase because airway pressure is high for a prolonged period. The presence of a leak may introduce errors in measurements of respiratory mechanics made at the airway opening. If so, respiratory mechanics can be measured more accurately when data are collected during the expiratory phase of ventilation. We examined this phenomenon in a lung model. When a leak was introduced into the model, simulating a leak around the TT, the leak occurred predominantly during the inspiratory phase of respiration. As the magnitude of the leak increased, the overestimation of resistance progressively increased, when calculated from pressure and flow measured at the airway opening. A large leak (38%) resulted in an overestimation of respiratory system resistance by 51% and an underestimation of elastance (Ers) by 23% when calculated from the entire ventilatory cycle. However, there was no under- or overestimation in Rrs when calculated from the expiratory phase only, and ERS was overestimated by only 6.1%. Varying peak inspiratory pressure, end-expiratory pressure, and expiratory time did influence the effect of leak, however, respiratory mechanics could still be calculated accurately from the expiratory phase under these conditions. We conclude that measurements of lung mechanics from the expiratory phase is a promising approach to dealing with the problem of measuring respiratory mechanics in mechanically ventilated infants with leaks around the tracheal tube.