In order to evaluate the influence of hypoxia on upper airway patency, we measured the response of upper airway resistance (UAR) to progressive (P) normocapnic hypoxia (Rebuck's method) and transient (T) hypoxia (three to five breaths to 100% N2) in 11 normal men. Breath-by-breath inspiratory UAR was calculated at isoflow during exclusive nasal breathing. The UAR response to hypoxia was characterized by the changes in nasal resistance and pharyngeal resistance (PR) as a function of Sa(O2), mean inspiratory flow (VT/TI), and changes in the end-expiratory lung volume (EELV) measured with an inductance vest. The ventilatory response to hypoxia was greater during T (-0.31 +/- 0.03 L/min/%Sa(O2); mean +/- SEM) than during P (-0.27 +/- 0.03 L/min/%Sa(O2), p = 0.05). UAR decreased as Sa(O2) decreased; this decrease was steeper during T than during P hypoxia (DELTA-PR/%Sa(O2): 3.9 +/- 0.5 during P and 2.5 +/- 0.2 during T, p = 0.05). For the whole group, there was no difference in the slope of the decrease in UAR with increasing VT/TI between the two hypoxic tests (DELTA-PR/DELTA-VT/TI: -0.85 +/- 0.1 during P and -0.70 +/- 0.1 during T, p > 0.05). However, in four subjects, the slope of the relationship PR/VT/TI during T remained steeper than during P. EELV increased as Sa(O2) decreased, with a greater increase during progressive than during transient hypoxia. We conclude (1) that the pharyngeal resistance response to hypoxia is greater during T than during P hypoxia, and (2) that these different responses are possibly due to differences in the central respiratory output (ventilatory) response to hypoxia and not to the differences in the hypoxia-induced pulmonary inflation.