A previously developed mathematical model that describes the relationship between blood alcohol (ethanol) concentration and the concentration of alcohol in the exhaled breath at end-exhalation (BrAC) has been used to quantitate the effect of pretest breathing conditions on BrAC. The model was first used to "condition" the airways with different breathing maneuvers prior to simulating a single exhalation maneuver, the maneuver used in standard breath alcohol testing. On inspiration, the alcohol in the air reaches local equilibrium with the alcohol in the bronchial capillary bed prior to entering the alveolar region. On expiration, approximately 50% of the alcohol absorbed on inspiration is desorbed back to the airways. BrAC correlates with the amount of alcohol that is desorbed to the airways. The six pretest breathing conditions and the percent change in BrAC relative to the control maneuver were: hyperventilation (-4.4%), hypoventilation (3.7%), hot-humid air (-2.9%), hot-dry air (0.66%), cold-humid air (0.13%), and cold-dry air (0.53%). The mechanism underlying these responses is not due to changes in breath temperature, but, rather to changes in the axial profile of alcohol content in the mucous lining of the airways.