The treatment and management of intensive care unit (ICU) patients requires the interpretation of a wide range of clinical data and measurements to decide the ideal mechanical ventilation (MV) settings. Often, ICU patients present with multiple organ dysfunctions so any interventions have to take into account the possible consequences on downstream organs; this is usually not possible due to inadequacies in monitoring and due to the complexity of patients. The result is that improvements in ICU patients can be inconsistent and unpredictable on a patient-to-patient basis. In worst case scenarios patients will receive MV for prolonged periods, which increase the risk of ventilator-induced lung injury (VILI). One key hurdle that is preventing progress in this field is the limited ability to identify precisely how a specific MV intervention will alter the working haemodynamic. Computational modelling can offer a real alternative to test separate hypothesis in a safe and effective manner; we are currently developing a computational simulation research tool that recreates, in great detail, the physiology of diseased patients receiving MV. Previous results have shown that the simulation is capable of providing realistic predictions of ideal MV settings [1]. In this study, our aim was to simulate ICU patient parameters to directly compare the model predictions with data reported in the literature. A multi component cardiovascular, respiratory and circulatory model has been integrated with the existing and previously validated pulmonary model developed by our lab [1-3]. The physiological interactions of the systemic and pulmonary circulation have been devised to allow for the study of multiple parameters including the heart rate, arterial pressure, cardiac output, ventilation/perfusion matching-pattern of the lung, and pressure-volume behaviour of the four cardiac chambers. The model has been validated by simulating low flow states with reduced oxygen delivery, common in ICU patients, to test the capabilities of the working model. The effects of varying parameters in one compartment of the model can be viewed in the global model to make accurate predictions about in silco patients. Computational models of physiological processes offer a viable alternative to expensive and time-consuming in vivo experiments, and offer the possibility of fixing variables to allow better observation of the effect of variables of interest -a feature not possible in clinical studies. 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. Our results are an important step forward in advancing the study of MV strategies and recruitment manoeuvres.