Drug-induced sleep endoscopy (DISE) is used to guide therapeutic management of obstructive sleep apnea (OSA), depending on the levels and patterns of pharyngeal collapse. However, the collapsibility of specific pharyngeal sites remains unknown. To assess collapse sites in patients with OSA undergoing DISE and whether number and location are associated with differences in airway collapsibility; and to quantify differences in collapsibility between primary and secondary sites in multilevel collapse. This cohort study assessed adult patients (≥18 years) with OSA undergoing DISE with manometry and positive airway pressure (PAP) titration at a tertiary care center from November 2021 to November 2023. Patients with an AHI score greater than 5 were included; those with less than 1 apnea event during DISE or incorrect catheter placement were excluded. Data were analyzed from September 28, 2022, to March 31, 2024. DISE with manometry and PAP titration. Active pharyngeal critical pressure (Pcrit-A) and pharyngeal opening pressure (PhOP) were used to quantify airway collapsibility, adjusted for covariates (age, sex, race, and body mass index [BMI]). Of 94 screened, 66 patients (mean [SD] age, 57.4 [14.3] years; BMI, 29.2 [3.9]; 51 [77.3%] males) with a mean (SD) apnea-hypopnea index (AHI) of 31.6 (19.0) were included in the analysis. Forty-seven patients (71.2%) had multilevel collapse, 10 (15.2%) had single-level nasopalatal collapse, and 9 (13.6%) had single-level infrapalatal collapse. Groups did not differ in demographic characteristics or established measures of OSA severity. The single-level nasopalatal group had substantially elevated levels of airway collapsibility (Pcrit-A and PhOP covariate adjusted mean, 2.4; 95% CI, 1.1 to 3.8; and 8.2; 95% CI, 6.4 to 9.9 cmH2O) compared to the single-level infrapalatal group (-0.9; 95% CI, -2.4 to 0.5 cmH2O; and 4.9; 95% CI, 3.0 to 6.8 cmH2O, respectively) and similar to the level among the multilevel group (1.3; 95% CI, 0.7 to 2.0; and 8.5; 95% CI, 7.7 to 9.3 cmH2O). The multilevel group had more negative inspiratory pressure (-24.2; 95% CI, -28.1 to -20.2 cmH2O) compared to the single-level nasopalatal group (-9.8; 95% CI, -18.3 to -1.28 cmH2O). In patients with multilevel collapse, airway collapsibility was significantly higher at the primary nasopalatal compared to secondary infrapalatal site (mean difference, 13.7; 95% CI, 11.3 to 16.1 cmH2O). The findings of this cohort study suggest that intervention should target the primary site of pharyngeal collapse, and secondary sites only if they are nearly as collapsible as the primary site. Future work is needed to precisely define the difference in primary and secondary collapsibility that necessitates multilevel treatment.