A retrospective analysis of positional data from 100 male patients with obstructive sleep apnea (OSA) was conducted to determine whether or not 1) the degree of positional dependency was similar in rapid eye movement (REM) compared to non-REM (NREM) sleep, 2) positional dependency correlated with effective levels of nasal continuous positive airway pressure (CPAP) and 3) patients with positional OSA preferentially avoided sleeping in the supine position. The apnea-hypopnea index (AHI) was scored separately for sleep state (NREM and REM) and for posture [off back (AHI-O) and on back (AHI-B)]. The ratio of AHI-O/AHI-B was used to define positional OSA as AHI-O/AHI-B less than or equal to 0.50 (P group) and nonpositional OSA as 0.50 less than AHI-O/AHI-B (NP group). A group of 31 patients who had sufficient sleep time in NREM and REM sleep in both sleep postures was selected. In this group 9 out of 22 subjects who showed positional dependency during NREM sleep became nonpositional during REM sleep (0.05 less than p less than 0.10). The mean effective nasal CPAP level was slightly, but significantly, lower in the P group than in the NP group (8.0 versus 9.1 cm H2O; p less than 0.05). In addition, a correlation between AHI and effective nasal CPAP levels was found (r = 0.491; p = 0.0001). The P group had less supine sleep time (SST) than the NP group (32% versus 45% of total sleep; p less than 0.005).(ABSTRACT TRUNCATED AT 250 WORDS)