Poor sleep and attention deficits are common in COPD. To assess the relationship between self-reported poor sleep and attention deficits in COPD. We also studied the association between self-reported sleep and the attention tests with the objective characteristics of sleep. Fifty-nine COPD patients were prospectively studied. Self-reported sleep quality was assessed using the Pittsburgh sleep quality index (PSQI). Objective characteristics of sleep were assessed by actigraphy and polysomnography. Attention was evaluated with the Oxford sleep resistance test (OSLER) and the Psychomotor vigilance test (PVT). 28 (47 %) patients referred poor sleep (PSQI >5). In the OSLER test they showed earlier sleep onset than patients with good sleep, median (Interquartil range): 31.2 min (25.4–40) vs 40 min (28.5–40), p: 0.048. They also spent more time making errors: 4.5 % (0.6–7.6) of total test time vs 0.7 % (0.2–5.3), p: 0.048. In PVT, patients with poor sleep presented a greater dispersion of the reaction time values with a higher value in the slowest 10 % of the reactions, 828 (609–1667) msec. vs 708 (601–993) msec, p: 0.028. No association was found between self-reported poor sleep and objective sleep variables. We found no correlation between OSLER and PVT results and polysomnographic variables except between sleep efficiency and PVT response speed (β: 0.309, p: 0.018). Self-reported poor sleep in COPD is associated with attention deficits. Sleep quality should be included in future studies of this facet of cognition in COPD, as well as to assess its potential usefulness as a therapeutic target. •Poor sleep quality is frequent in COPD patients.•Self-reported poor sleep is associated with attention deficits in COPD patients.•COPD patients with reported poor sleep make more errors and have slower reactions.•Pittsburgh questionnaire and polysomnographic variables are not related in COPD.