Objectives We test the hypothesis that the incidence of sleep problems are influenced by sociodemographic variables, particularly social deprivation, and ethnicity.

Methods Self-reports of sleep duration and sleep difficulties (waking in the night, sleeping in the day, difficulty waking and snoring), personal wealth (household income, property owning etc), ethnic group, employment, education, as well as post-code-based Townsend Social Deprivation, were extracted from UK Biobank’s cohort of c500,000 British-domiciled adults (40–70 years). Analyses contrasted the incidence of different sleep problems, and a composite measure of these (the Problematic Sleep Index) - across groups.

Results Almost one-third of participants reported sleeping shorter (24.7%) or longer (7.7%) than age-corrected recommended sleep durations. The incidence of shorter or longer sleep increased with social deprivation and varied with ethnicity. Snoring, waking during the night, finding it difficult to get up in the morning and sleeping in the daytime, were subject to similar effects. The Problematic Sleep Index showed being younger, male, employed, home-owning, higher household income, higher level of educational achievement and time in education, were all associated with better sleep, as was living in a more affluent area and being white.

Conclusions Sleep problems in Britain show a social gradient, independently of a range of other demographic and social influences, suggesting that sleep quality differs with and between ethnic groups. These sleep inequalities suggest that the protective and recuperative effects of sleep are disproportionately distributed across society and should encourage us to consider the potential benefits of community specific sleep interventions.