Abstract Introduction Obstructive sleep apnea (OSA) is characterized by repetitive episodes of pharyngeal obstruction at one or more collapse sites. Multilevel collapse is a negative predictive factor for PAP-alternative surgical treatments. Pharyngeal manometry during drug-induced sleep endoscopy (DISE) permits an objective evaluation of anatomical obstruction sites and mechanisms, and may play a role in patient selection for PAP alternative therapies. We hypothesized that patients with multilevel collapse would have primary velopharyngeal collapse and secondary, downstream sites. Methods This prospective cohort of twenty consecutive patients with diagnosed OSA underwent DISE with nasal airflow monitoring and pharyngeal manometry. Two pressure catheters were placed at the (1) distal margin of the soft palate (Pus) and (2) retroepiglottic space (Pds). We sought to delineate site(s) of collapse by evaluating dissociations in airflow and air pressure during inspiratory flow limitation and/or complete obstruction. The primary site of collapse was defined by the initial segment at which inspiratory airflow and pressure dissociated (the critical pressure, Pcrit). Secondary site(s) were defined as the pressure at which the Pus and/or Pds signals subsequently dissociated (beyond the initial point of maximal flow). Results Twenty-one patients were evaluated of which signal quality and catheter placement in twenty were sufficient for analysis. Participants were elderly (58.3±14.3yrs, mean±SD), overweight (BMI 28.7±2. kg/m2), and had moderate-severe OSA (AHI 28.2±14.2 events/hour). Six patients had a single site of collapse, of which one was naso-palatal (Pcrit -2.6cmH2O) and five were infra-palatal (mean Pcrit -1.3cmH2O). In the patients with multilevel collapse (N=14), all had primary naso-palatal collapse (mean Pcrit 0.17±2.0cmH2O) and secondary infra-palatal sites of collapse (mean Pcrit -11.7±5.8cmH2O). Pcrit differed between the primary and secondary site by 11.9±6.5cmH2O (p< 0.01). Conclusion These findings suggest that single-site pharyngeal collapse is usually infra-palatal. In contrast, multilevel collapse is primarily due to velopharyngeal collapse with varying degrees of instability in downstream pharyngeal segments. When secondary infra-palatal sites readily collapse with minimal additional respiratory effort, laxity of the tongue and/or lateral walls may be strongly implicated in the pathogenesis of pharyngeal obstruction. Future studies can utilize pharyngomanometry to further target specific pharyngeal structures for OSA therapy. Support (if any) National Institutes of Health 1R01HL144859 (PIs: Schwartz, Dedhia)