To examine the effects of sustained hypoxia on upper airway and chest wall muscle activity in humans, we measured genioglossus muscle (GG) activity, inspiratory intercostal muscle (IIM) activity, and ventilation during sustained hypoxia in 17 normal subjects and 17 patients with obstructive sleep apnea (OSA). The trial of sustained hypoxia was performed as follows: after an equilibration period of 3 min, isocapnic hypoxia (arterial O2 saturation = 80 +/- 2%) was maintained for 20 min. GG EMG was measured with a fine-wire electrode inserted percutaneously, and IIM EMG was measured with surface electrodes. Ventilatory response to sustained hypoxia was initially increased and subsequently decreased. Stable phasic GG activity during spontaneous tidal breathing was observed in 6 normal subjects and 10 patients with OSA. Responses of GG and IIM activities to sustained hypoxia showed a biphasic response qualitatively similar to the ventilatory response in these 16 subjects. The absolute value of the subsequent decline in GG activity was similar to that of the initial increase, whereas the subsequent decline in IIM activity was smaller than that of the initial increase. Percent GG activity was significantly lower than both percent IIM activity and percent minute ventilation during the decline and plateau phases. There were no significant differences in ventilatory and EMG responses between the normal subjects and the patients with OSA. We conclude that, during wakefulness, upper airway muscle activity declined to a greater extent than inspiratory pump muscle activity during sustained hypoxia.