Objective: To determine the site(s) of upper airway obstruction in patients with obstructive sleep apnea syndrome (OSAS) and their changes during sleep with upper airway (UA) pressure measurement; and to analyze the correspondence between the UA pressure measurement and polysomnography (PSG). Method: Thirty patients with OSAS underwent UA pressure measurement. The catheter with five solid-state ultraminiature sensors was inserted through the patients' upper airway to the esophagus. The sensors were located at the nasopharynx, oropharynx, tongue base, hypopharynx, and esophagus. The lower limit of UA obstruction was determined by relying on the observed pressure pattern. Result. 1) During inspiration, obstruction occurred associated with an increased negative inspiratory pressure inferior to the site of obstruction and a disappeared negative inspiratory pressure above the site of obstruction; 2) three patterns of obstruction were observed; 3) the site of obstruction was located at the site of the palate; 4) velopharyngeal obstruction and tongue-pharynx obstruction were all present; 5) nasopharyngeaI obstruction velopharyngeal obstruction and tongue-pharynx obstruction were all present; and 6) peak inspiratory pressure differences between the normal breath and apnea were associated with the longest apnea interval (P <.05). Conclusion: UA pressure measurement can evaluate the site of obstruction in patients with OSAS and their changes during sleep. Peak inspiratory pressure differences between normal breath and apnea can show the severity of OSAS.