Computational fluid dynamics (CFD) analysis may quantify the severity of anatomical airway restriction in obstructive sleep apnea syndrome (OSAS) better than anatomical measurements alone. However, optimal CFD model endpoints to characterize or assess OSAS have not been determined. To model upper airway fluid dynamics using CFD and investigate the strength of correlation between various CFD endpoints, anatomical endpoints, and OSAS severity, in obese children with OSAS and controls. CFD models derived from magnetic resonance images were solved at subject-specific peak tidal inspiratory flow; pressure at the choanae was set by nasal resistance. Model endpoints included airway wall minimum pressure (P_min), flow resistance in the pharynx (R_pharynx), and pressure drop from choanae to a minimum cross section where tonsils and adenoids constrict the pharynx ( dP _TAmax). Significance of endpoints was analyzed using paired comparisons ( t-test or Wilcoxon signed rank test) and Spearman correlation. Fifteen subject pairs were analyzed. R_pharynx and dP _TAmax were higher in OSAS than control and most significantly correlated to obstructive apnea-hypopnea index (oAHI), r = 0.48 and r = 0.49, respectively ( P < 0.01). Airway minimum cross-sectional correlation to oAHI was weaker ( r = −0.39); P_min was not significantly correlated. CFD model endpoints based on pressure drops in the pharynx were more closely associated with the presence and severity of OSAS than pressures including nasal resistance, or anatomical endpoints. This study supports the usefulness of CFD to characterize anatomical restriction of the pharynx and as an additional tool to evaluate subjects with OSAS.