Acute lung injury (ALI) defined by hypoxemia (PaO2/FiO2 ≤300mmHg) and its severe form acute respiratory distress syndrome (ARDS) (PaO2/FiO2 ≤200mmHg) account for 79 and 59 cases per 100,000 persons respectively [1]. Such patients require mechanical ventilation (MV) and the optimal ventilator settings are often difficult to determine for individual patients and can further induce stretch related lung injury. The ARDSNET trial [2] crucially showed that lowering tidal volume from 12 to 6mL/kg had a significant reduction in mortality and time spent on MV. Following this publication little progress has been made in elucidated the ideal settings for MV such as PEEP and application of recruitment manoeuvres (RM) to further lower mortality which is often difficult to study in the clinical setting. Computational modelling can offer a real alternative to test separate hypothesis in a safe and effective manner and our aim was to simulate ARDS disease patient parameters to directly compare the model predictions with data reported in the ARDSNET trial. The ARDSNET patient database was accessed with permission NHLBI Biologic Specimen and Data Respiratory Information Centre to create virtual ARDS patients and analysed using the Nottingham Physiological simulator (NPS) [3]. The baseline MV parameters including tidal volume, arterial blood gases, PEEP and airway pressures used in the ARDS Network study were computed into the NPS to analyse the outcomes. Previous results have shown that the simulator is capable of providing realistic predictions of ideal MV settings [4] and work is currently underway to increase the number of virtual patients used to compare actual and predicted outcomes from mechanical ventilation. Mathematical models of physiological processes offer a real alternative to expensive and time consuming in vivo experiments as a single variable can be studied in isolation. The development of such tools will offer clinicians a more detailed insight into the disease process to help them determine the best strategy for improving oxygenation. However such models need to be rigorously tested to account for the varying uncertainty in the patient populations. The ARDSNET study contained data from 861 ARDS patients and will prove invaluable in the further development and validation of the NPS.