Rationale: There have been meta-analyses that showed reduced retinal nerve fiber layer (RNFL) thickness, which is a surrogate marker of glaucoma, in patients with obstructive sleep apnea (OSA). However, the sample sizes in these reports were small (<300), and the mechanism of RNFL thinning in patients with OSA was not revealed.Objectives: To investigate the relationship of RNFL thickness with nocturnal hypoxemia or hypoxemic burden in a large-scale study.Methods: In this epidemiological study, 8,309 community residents were enrolled. The actigraphy-modified 3% oxygen desaturation index (acti-ODI3%) and cumulative percentage of sleep time with oxygen saturation <90% (acti-CT90) modified by objective sleep duration using actigraphy were measured. The hypoxemic burden is shown as acti-CT90. Circumpapillary RNFL thickness was determined using optical coherence tomography.Results: Multivariable logistic analysis models revealed that an increase in acti-CT90 was significantly associated with mean RNFL thinning after adjusting for several factors in participants without glaucoma diagnosed or treated previously (β = -0.037; P = 0.009). There were significant differences in mean RNFL thickness among participants stratified according to acti-CT90 (>1.5 vs. ⩽1.5; P = 0.04). Although acti-ODI3% was significantly associated with acti-CT90 (β = 0.72; P < 0.0001), acti-ODI3% was not significantly associated with mean RNFL thickness in the multivariable logistic analysis (β = -0.011; P = 0.48). In addition, acti-CT90 was significantly associated with mean RNFL thickness both in the elderly (⩾60 yr; β = -0.058; P = 0.002) and nonelderly (<60 yr; β = -0.054; P = 0.007).Conclusions: Acti-CT90, but not acti-ODI3%, was associated with mean RNFL thinning in participants irrespective of age in the elderly or nonelderly. Further prospective studies are required to investigate whether the prevention of hypoxic burden, which was shown as acti-CT90 in this study, is favorable for RNFL thinning.