Mechanical Ventilation (MV) therapy is one of the most common treatments offered to patients with respiratory failure in ICU. MV assists patient recovery by completely or partially taking over the breathing process and helping with oxygen delivery and removal of carbon dioxide. However, inappropriate MV settings mismatched to a given patient’s condition can cause further damage. On the other hand, suboptimal MV settings can increase the length of stay of the patient in ICU and increase the cost of treatment. Acute Respiratory Distress Syndrome (ARDS) is a major form of Acute Lung Injury (ALI) where clinicians offer a supportive environment for patient recovery by application of MV. ARDS is characterised by inflamed and fluid filled lungs that result in alveolar collapse and thus severe hypoxemia. Application of positive end expiratory pressure (PEEP) is employed to recruit and retain lung units to maximise gas exchange. However, a delicate trade-off is required between maximising gas exchange and preventing further unintended damage to the lungs, when determining optimum PEEP level. Currently, no specific protocols to determine optimum PEEP level exist and selection of PEEP is dependent on medical intuition and experience, primarily due to lack of easy methods to determine patient – specific condition at the patient’s bedside. A mathematical recruitment model is developed in Labview to help determine patient – specific condition based on fundamental lung physiology and engineering principals in this thesis. The model utilises readily available clinical data to determine parameters that identify underlying patient – specific lung characteristics and conditions. Changes in these parameters can be monitored over time and compared between patients to determine the severity of the disease and evolution of disease with time. A second model is developed to determine dynamic functional residual capacity (dFRC), that represents the extra volume retained in a lung through application of PEEP. The model extends previous efforts in the field that applied the stress – strain theory to lung mechanics to estimate dFRC. This model estimates the patient’s dFRC using readily available clinical data (PV data) and can be monitored over time to determine changes in a xiii given patient’s condition. The dFRC model introduces a new parameter, , which is considered a population constant for the particular PEEP. The model offers an easy and reliable method to determine dFRC since other methods are normally invasive or require interruption of MV. The models developed were validated against real – time clinical data obtained through clinical trials. The recruitment model was found to fit the clinical data well with error values within acceptable limits. It also enabled identification of parameters that reflect the underlying patient – specific lung condition. The dFRC model was able to estimate the dFRC for a patient with high level of accuracy for clinically applicable PEEP levels. The two models work well in conjunction with each other and provide a novel and easy method to clinicians to determine patient – specific lung characteristics and ultimately determine optimal MV treatment parameters, especially PEEP.