Epidemiological studies have identified an association between obstructive sleep apnea (OSA) and systemic hypertension, but no experimental studies have addressed this relationship nor established that OSA in fact causes sustained daytime hypertension. Accordingly we examined the effects of OSA on blood pressure (BP) in a canine model of induced OSA. Sleep-wake state of the dog was monitored overnight by a computer that received telemetered signals from permanently implanted EEG and EMG electrodes. Whenever sleep was detected, the computer sent a signal to close a valve connected to the endotracheal tube through which the dog breathed, thereby producing airway occlusion; when the dog aroused from sleep, the occlusion was released. Thus the sequence of events simulated closely human OSA. BP was measured continuously by a permanently implanted arterial catheter and telemetry system. Each of four dogs was studied before, during, and after a period of OSA of 1 -4 months' duration, ïn each dog, OSA (of 50-60 apneas/hr of sleep) produced an increase in mean nighttime BP (of 13+2 mm Hg, mean+se; p<0.001 ); and resulted in sustained daytime hypertension (increase in mean BP, 14+4 mm Hg; p<0.01). Following recovery each dog then underwent a protocol to produce recurrent arousals from sleep, without airway occlusion, by having the computer activate an acoustic alarm, rather than occluding the endotracheal tube. The arousal frequency in this protocol was very similar to that produced by OSA. In contrast to OS A, sleep fragmentation by recurrent arousals produced considerably less daytime hypertension than did OSA (p<0.001). The findings demonstrate that OSA in the dog results in sustained daytime hypertension that cannot be attributed solely to the sleep fragmentation that is associated with the sleep-apnea syndrome.