Findings from manikin-based studies on ventilation are commonly directly extrapolated to clinical practice. The aim of this study was to determine how the use of manikins affects measurements of ventilatory variables. We connected a lung simulator to a manikin, which was then ventilated at different inspiratory flows. We defined three experimental models to compare measurements of ventilatory variables between the mechanical ventilator and the lung simulator. Even when no leakage occurred, significant tidal volume deviations were observed; from a mean (SD) of 21 (2) ml to 49 (9) ml, and from 40 (4) ml to 88 (5) ml for invasive and non-invasive ventilation, respectively (p < 0.001). Significant peak pressure deviations from 0.7 (0.1) cmH2O to 10.6 (0.3) cmH2O were also recorded during non-invasive ventilation (p < 0.001). Evaluation of manikin resistance and airway dead space may be essential to limit study bias. We suggest a recalibration of the recorded data if comparisons are made between different tests performed at different inspiratory flows.