Current evidence suggests that patients with obstructive sleep apnea (OSA) may have augmented pharyngeal dilator muscle activity during wakefulness, to compensate for deficient anatomy. However, the isolated effect of sleep on the activity of these muscles (comparing OSA patients with controls) has not been studied. We therefore determined waking levels of genioglossus (GG) and tensor palatini (TP) muscle activity (% of maximum electromyographic [EMG] activity) in 10 OSA patients and eight controls, and then assessed the impact of the first two breaths of sleep (theta electroencephalographic [EEG] activity) following a period of stable wakefulness. Apnea patients demonstrated greater genioglossal (27.4 +/- 4.0 versus 10.7 +/- 2.1%) and tensor palatini (31.9 +/- 6.5 versus 10.6 +/- 1.9%) EMG activity than did controls during wakefulness. This augmented muscle activity in apnea patients could be reduced to near control levels during wakefulness with the application of continuous positive airway pressure (CPAP) to the upper airway. At sleep onset, control subjects demonstrated small but consistent decrements in the activity of both the TP and GG muscles. On the other hand, apnea patients demonstrated large, significantly greater decrements in TP EMG at sleep onset than did the control subjects. The effect of sleep on GG EMG in apnea patients was inconsistent, with most (n = 7) demonstrating large (significantly larger than controls) decrements in genioglossal activity. However, three OSA patients demonstrated small increments in GG EMG at sleep onset despite falling TP EMG and obstructive apnea or hypopnea. We conclude that sleep onset is associated with significantly larger decrements in TP muscle EMG activity in OSA patients than in controls, which may represent a loss of neuromuscular compensation that is present during wakefulness. However, our results for the GG muscle were more variable, and did not always support this hypothesis.