Nasal high flow (NHF) therapy is a recent form of non-invasive respiratory support for patients suffering from respiratory distress that supplies high flows of heated and humidified air, oxygen or a mix via a nasal cannula. A number of in vivo studies have proven its effectiveness at improving blood oxygenation; however, its mechanisms of action remain widely unproven. Two proposed mechanisms of action, the CO2 washout of anatomic dead space and the production of positive airway pressure, are investigated in this thesis for the use of the Fisher & Paykel Healthcare Ltd (FPH) Optiflow™ adult nasal cannula through a range of experiments. Five anatomically correct upper airway models produced from computed tomography (CT) scan data via 3D printing were employed during in vitro experiments and two live subjects participated in in vivo measurements. The human respiratory system was faithfully replicated for CO2 washout experiments with physiological CO2 diffusion into the lung replicated by a constant flow of CO2 into the lung pump. In vivo measurement of a natural breathing flow pattern was scaled to an average population tidal volume and respiratory rate for in vitro use. In vitro measurements of static pressure during natural breathing found similar flow resistances across the nasal passage for inspiratory and expiratory flow directions; however, across the entire upper airway greater resistance was seen for inspiration. Introduction of NHF therapy produced significant increases in all mean and peak airway pressures within the upper airway with a flow rate of 30 LPM fulfilling the inspiratory work requirements presented by the upper airway resistance. In vivo and in vitro hot wire anemometry measurements at the exterior nares indicated low velocity and turbulence intensity flows at peak inspiration and a high velocity jet with high turbulence during peak expiration. At natural breathing an in vitro anterior-posterior velopharynx traverse captured low turbulence intensities during peak inspiration and high turbulence intensities during peak expiration. Introduction of NHF therapy had little influence on the turbulence intensity profile of peak expiration yet did cause significant increases in the turbulence intensities during peak inspiration. Measurements of the CO2 concentration near the lung volume over many breath cycles were used to find time-averaged CO2 concentrations. For the standard airway model an average CO2 concentration of 4.88 ± 0.07 %V/V was determined during natural breathing. Implementation of increasing levels of NHF therapy generated significant washout of CO2 reducing this average concentration to a minimum of 3.81 ± 0.11 %V/V at a flow rate of 80 LPM. It was determined that airway geometry significantly affected the efficacy of the NHF therapy though CO2 washout was observed in all five airway models.