A hypoxic ramp test (to 75% arterial oxygen saturation) during the awake state was performed in 17 patients with obstructive sleep apnea (OSA). Blood pressure was monitored with an indwelling arterial line (radial artery), and the ventilatory response to eucapnic hypoxia was determined. Eight of the patients were normotensive. The remaining nine hypertensive patients were studied after a 3-wk washout of antihypertensive medication. Compared with a nonmatched group of normotensive nonsnoring control subjects in whom hypoxemia did not affect blood pressure, all OSA patients showed various degrees of pressor responses during hypoxia. The pressor response was of similar magnitude in normotensive and hypertensive patients with OSA. There was a significant relationship between the ventilatory and the pressor responses to hypoxia (p = 0.03) that was similar in both normotensive and hypertensive patients. Although disease severity expressed as apnea index (number of apneas per hour of sleep) or minimum arterial oxygen saturation reached during the overnight recording correlated with the magnitude of the pressor response (p = 0.03 and 0.045, respectively), the ventilatory response to hypoxia was unrelated to disease severity. Hypoxemia induced a similar increase in heart rate in controls and in normotensive patients, but an attenuated heart rate response was seen at the nadir of oxygen saturation in hypertensive subjects (p < 0.05). These data demonstrate that patients with obstructive sleep apneas have a pressor response to hypoxia. This response is likely to be involved in the blood pressure swings seen during apnea in patients with OSA.(ABSTRACT TRUNCATED AT 250 WORDS)