To evaluate the incidence of sleep apnea syndrome (SAS), oxygen desaturation during sleep and sleep quality in patients with chronic obstructive pulmonary disease (COPD), 30 COPD patients and 20 healthy snorers (without SAS) were studied. Each subject received a pulmonary function test (PFT) (simple spirometry), arterial blood gas determination and an overnight sleep study. COPD patients were divided into two groups: those with SAS (group I) and those without (group II). Group II patients were further subdivided into: group IIa [delta SaO2 < 15% (delta SaO2 = baseline SaO2-lowest SaO2) and group IIb (delta SaO2 > or = 15%); group IIc (baseline SaO2 > 90%) and group IId (baseline SaO2 < or = 90%). Our results showed that only six of 30 (20%) COPD patients had an associated SAS. Among group II patients, the lowest SaO2 and delta SaO2 were correlated with baseline SaO2, PaO2 and PaCO2 but were not correlated with age, % of ideal body weight, FVC, FEVl, FEVl/FVC. Group IIb patients (n = 10) had a lower SaO2 during sleep, a lower baseline PaO2, and a lower hematocrit level than group IIa patients (n = 14). Group IId patients (n = 9) had a lower PaO2 and a higher delta SaO2 during sleep than group IIc patients (n = 15). However, there were no significant differences in age, percent of ideal body weight (IBW), FVC, FEVl or FEVl/FVC values between groups IIa and IIb, or between groups IIc and IId. Group II patients had a lower percentage of sleep efficiency, higher arousal and movement indices and a longer period of stage 1 sleep, compared to the control group. In conclusion, the incidence of COPD patients with SAS is 20%. Age, percentage of IBW, FEVl, and FEVl/FVC values are not reliable predictors of the risk of nocturnal hypoxemia among COPD patients. However, a possible correlation between baseline SaO2, PaO2 and PaCO2 values and the incidence of nocturnal hypoxemia exists. Finally, COPD patients experienced a poorer quality of sleep in comparison with the control group.