During hypoxia ATP degradation to uric acid is increased in animal models and humans. To assess the reliability of an overnight increase in uric acid excretion as a marker of nocturnal hypoxemia, we selected 10 normal volunteers (7 males and 3 females), 29 COPD patients (26 males and 3 females), and 49 subjects with obstructive sleep apnea (OSA) (43 males and 6 females). The patients underwent standard polysomnography, which was repeated in 14 subjects with nasal continuous positive airway pressure (CPAP), and were subdivided into two groups: Group D included desaturating subjects who spent at least 1 h at SaO2 < 90% and 15 min below 85%, and Group ND were nondesaturating subjects. The overnight change in the uric acid:creatinine ratio (delta UA:Cr) was negative in normal subjects (-27.5 +/- 9.1 [mean +/- SD]) and ND groups: -19.7 +/- 14.3 in COPD, -16.1 +/- 13.0 in OSA. In both COPD and OSA Group D, the ratio was usually positive: delta UA:Cr was 17.9 +/- 31.4 in Group D COPD (p < 0.001 versus ND) and 10.1 +/- 30.7 in Group D OSA (p < 0.001 versus ND and versus normal subjects) despite 4 of 15 false negative results in COPD and 8 of 20 in OSA. CPAP effective treatment induced a marked reduction ((p = 0.0024) in delta UA:Cr, leading to a negative value. We conclude that delta UA:Cr seems to be a promising index of significant nocturnal tissue hypoxia, with good specificity but poor sensitivity (about 30% false negative), which might be useful for the long-term follow-up of outpatients on nasal CPAP with a positive ratio at baseline.