Introduction Accurate measurements of ventilatory outputs are crucial during mechanical ventilation support. These measurements are achieved through sensors that monitor parameters such as flow/volumes and pressures. External and internal flow sensors are both commonly used in mechanical ventilation systems to measure the flow of air entering and leaving the patient’s lungs. The sensors could be located outside the ventilator (external or proximal) or inside the ventilator (internal or distal), each of which have their own respective advantages and disadvantages. There are differences in the way they function and the information they provide, which can affect their accuracy and usefulness in different clinical situations. A few clinical studies have compared the use of external to internal sensors in mechanical ventilation, showing mixed results. The intent of this study is to reexamine the differences between two critical care ventilators utilizing external sensors to two other ventilators utilizing internal sensors. Methods A bench study using a lung simulator was conducted. We constructed three passive, single compartment models: 1) compliance of 40 ml/cmH2O, resistance of 10 cmH2O, 2) compliance of 40 ml/cmH2O, resistance of 20 cmH2O, 3) compliance of 20 ml/cmH2O, resistance of 10 cmH2O. In each experiment we used two different modes of ventilation, volume controlled (tidal volume 400 ml, respiratory rate 20, PEEP 5 cmH2O, inspiratory time 0.7 seconds) and pressure controlled (inspiratory pressure 15 cmH2O, respiratory rate 20, PEEP 5cmH2O, inspiratory time 0.7 seconds). We compared the inspiratory flow, inspiratory tidal volume, peak inspiratory pressures and PEEP in four commercially available critical care ventilators. Two use external flow sensors: G5 (Hamilton Medical), Bellavista 1000e (Vyaire Medical), and two use internal flow sensors: Evita Infinity 500 (Drager), and PB 980 (Medtronic). We also compared these parameters to a mathematical model. Statistics to compare the parameters in all four ventilators were done with Kruskal Wallis test followed by post-hoc Dunn’s test. A t-test was used to compare the parameters between the ventilator-measured and simulator-measured parameters in each ventilator. A P value of < 0.05 was considered significant. Results There were statistically significant differences (P < 0.001) in all four measured parameters: inspiratory flow, tidal volume, PIP and PEEP between all four ventilators, and between the mathematical model and all four ventilators in both modes, in all three clinical scenarios. The post-hoc Dunn test showed significant differences between each ventilator, except for a few parameters in PIP and PEEP. but not in flow or volume. There were variable but significant differences between some of the four parameters measured from the ventilator compared to those measured from the simulator of all four ventilators in both modes. The two ventilators using external sensors had more accurate differences between the delivered and measured tidal volumes (P < 0.001) and inspiratory flow (P < 0.001), however, the other two ventilators with internal sensors had more accurate differences between the delivered and measured PIP (P < 0.001) and PEEP (P < 0.001) levels. Conclusions All four ventilators performed differently from each other and from the mathematical model. The two ventilators using external sensors had more accurate differences between the delivered and measured tidal volumes and inspiratory flow, the two ventilators with internal sensors had more accurate differences between the delivered and measured PIP and PEEP levels. Differences between the ventilators depends on multiple factors including location, type of sensor, and respiratory mechanics. Keywords Flow sensor, Pressure sensor, PIP, PEEP, Tidal volume, Flow