Obstructive sleep apnea syndrome (OSAS) is a disorder that causes sleep apnea and hypopnea, which in turn causes various disorders in daily life. Because of the difficulty in measuring airflow dynamics, computational fluid dynamics (CFD) simulations are performed to evaluate upper airway airflow in OSAS in detail. However, the relationship between the severity of OSAS, as measured by the apnea hypopnea index (AHI), and airflow dynamics is unclear. In this study, CFD simulations of human snoring during sleep were performed to determine the correlation between AHI and pressure drops in the nasal cavity and throat, as well as between AHI and minimum cross-sectional area of the throat. For the simulation, 3D models of snoring in the open-mouth state, which is a common form of snoring, were reconstructed based on computed tomography images acquired from four patients with mild OSAS and six with severe OSAS. Each relationship was evaluated using Spearman’s rank correlation coefficient. The correlation coefficient between AHI and pressure drop in the nasal cavity was 0.745, with a significant correlation. There was no significant correlation between AHI and pressure drop in the throat or between AHI and minimum cross-sectional area of the throat. These results suggest that the pressure drop in the nasal cavity affects the severity of OSAS.