Current evidence suggests that patients with obstructive sleep apnea (OSA) may have greater pharyngeal critical pressure (Pcrit), which reflects the increase in upper airway collapsibility. The contribution of Pcrit to the severity of OSA and to the efficacious continuous positive pressure (nCPAPeff) therapy has never been extensively described and no data are available about the interaction of Pcrit, age, and anthropometric variables. To determine the relationship between Pcrit, severity of the disease, nCPAPeff, and anthropometric variables we measured Pcrit in a group of 106 patients with OSA. Pharyngeal critical pressure was derived from the relationship between maximal inspiratory flow and nasal pressure, Pcrit representing the extrapolated pressure at zero flow. Upper airway resistance (Rus) was determined as the reciprocal of the slope (DeltaPn/DeltaVImax cm H2O/L/s) in the regression equation. In a subgroup of 68 patients, during the diagnostic night, we measured as indices of respiratory effort, the maximal inspiratory esophageal pressure (Pes) at the end of apnea (Pesmax), the overall increase from the minimum to the maximum (DeltaPes), and the rate of increase of Pes during apnea (RPes). As a group, the mean Pcrit was 2.09 +/- 0.1 cm H2O (range, 0 to 4.5) and the mean Rus was 11.1 +/- 0.5 cm H2O/L/s. Although men have greater Pcrit, pharyngeal collapsibility was influenced neither by neck size nor by body mass index (BMI). Although there was a significant relationship between Pcrit and apnea plus hypopnea index (AHI) (r = 0.23, p = 0.02), neck circumference was the stronger predictor of apnea frequency, with Pcrit contributing only to the 3% of the variance. In the group of patients as a whole, a model including AHI, BMI, Rus, and Pcrit explained the 36% of the variance in nCPAPeff, with a greater contribution of AHI, Pcrit accounting for only 3% of the variation. In patients for whom the measure of respiratory effort was obtained, 42% of the variance in nCPAPeff was explained by RPes (33%) and BMI. From these results we conclude that Pcrit alone does not yield a diagnostically accurate estimation of OSA severity and nCPAPeff. Although individual collapsibility may predispose to pharyngeal collapse, upper airway occlusion may require the combination of several factors, including obesity, upper airway structure, and abnormalities in muscle control.