The underlying pathophysiological mechanisms of exercise-induced laryngeal obstruction (EILO) remain to be fully established. It is hypothesised that high inspiratory flow rates can exert a force on laryngeal airway wall that contributeto its inward collapse causing obstruction. Computational fluid dynamics (CFD) presents an opportunity to explore the distribution of forces in a patient specific upper airway geometry. The current study combined exercise physiological data and CFD simulation to explore differences in airflow and force distribution between an EILO patient and a healthy-matched control. Subjects underwent incremental exercise testing with continuous recording of respiratory airflow and laryngoscopic video, followed by an MRI scan. The respiratory and MRI data were used to generate a subject specific CFD model of upper respiratory airflow. In the EILO patient, the posterior supraglottis experiences an inwardly directed net force, whose magnitude increases nonlinearly with larger flow rates, with slight changes in the direction toward the centre of the airway. The control demonstrated an outwardly directed force at all regions of the wall, with a magnitude that increases linearly with larger flow rates. A comparison is made between the CFD results and endoscopic visualisation of supraglottic collapse, and very good agreement is found. We present the first hybrid physiological / computational approach to investigate the pathophysiological mechanisms of EILO. The method shows great potential and the preliminary findings should be confirmed in larger sample sizes.