Objective. To assess the accuracy of questions evaluating the quality and duration of sleep. Material and methods. Target population included residents of St. Petersburg, who participated in the epidemiological study ESSE-RF. Out of 1417 participants, 136 individuals agreed to undergo additional sleep evaluation, and 49 subjects underwent polysomnography study for objective sleep evaluation. All participants were interviewed for subjective sleep assessment (standard questionnaires and Pittsburgh sleep quality index questionnaire). Results. One third of the participants evaluated sleep quality in the lab worse than at home. Their sleep duration was 1,5h shorter than in those who reported sleep quality to be as good as at home. However, the majority of participants slept in the sleep lab more than at home for the previous month. Moreover, 59% subjects reported sleep disturbances. Sleep duration during night in the lab was 6.6 [3; 9] — vs. 7 [3; 10] hours as assessed by PSG and by subjective report, respectively (p=0,06). At the same time, usual average sleep duration was 8 [5; 10] hours (p=0.005), and average sleep duration during last month — 7 [5; 9] hours. Conclusion. One third of subjects report worse sleep in the laboratory compared to the usual nights and it is associated with their subjective shorter sleep duration. Subjective assessment of sleep duration and sleep onset time is appropriate for sleep evaluation during the last night, unlike subjective assessment of awakenings after sleep onset. Average self-reported sleep duration depends on the question formulation.