Purpose: Diabetic retinopathy (DR) is a hypoxic retinal disease, but so far, the association with systemic hypoxia is poorly understood. Hence, the aim of this study was to evaluate cross-sectional and longitudinal associations between DR and chronic respiratory failure (CRF) in a national cohort. Design: Cross-sectional and 5-year longitudinal register-based cohort study. Methods: Between 2013 and 2018, we included patients with diabetes from the Danish Registry of Diabetic Retinopathy, who were each age and sex matched with five controls without diabetes. At index date, the prevalence of CRF was compared between cases and controls, and the longitudinal relationship between DR and CRF was assessed in a five-year follow-up. Results: At baseline, we identified 1,980 and 9,990 patients with CRF among 205,970 cases and 1,003,170 controls. The prevalence of CRF was higher among cases than controls (OR 1.75, 95% CI 1.65–1.86), but no difference between cases with and without DR was found. During follow-up, we identified 1,726 and 5,177 events of CRF among cases and controls, respectively. The incidence of CRF was higher among both cases with and without DR compared to controls (DR level 0: HR 1.24, 95% CI 1.16–1.33, DR level 1–4: HR 1.86, 95% CI 1.63–2.12), and higher among cases with DR compared to cases without DR (HR 1.54, 95% CI 1.38–1.72). Conclusion: In this study based on nationwide data, we found an increased risk of present and incident CRF in patients with diabetes with or without DR, and we identified DR as a predictor of future CRF.