To test the associations between sleep indices and eating behaviours in young adults, a group vulnerable to suboptimal sleep.

Cross-sectional analysis of survey measures of sleep (i.e. time in bed, variability, timing and quality) and dietary patterns (i.e. breakfast skipping, eating at fast-food restaurants, consumption of sports and energy drinks, and sugar-free, sugar-sweetened and caffeinated beverages).

Minneapolis/St. Paul metropolitan area of Minnesota (USA).

A total of 1854 respondents (20–30 years, 55·6 % female) from the 2008–2009 survey conducted for the third wave of the population-based Project EAT (Eating and Activity in Teens and Young Adults) study.

After adjustment for demographic and behavioural covariates in linear regression models, those who went to bed after 00.30 hours consumed 0·3 more servings of sugar-sweetened beverages per day, consumed 1·7 times more energy drinks, skipped breakfast 1·8 more times per week and consumed fast food 0·3 more times per week compared with those who went to bed before 22.30 hours. Reported sleep quality in the lowest (Q1) v. highest (Q3) tertile was associated with more intake of energy drinks (Q3 v. Q1, prevalence ratio, 95 % CI: 1·79, 1·24, 2·34), sports drinks (1·28, 1·00, 1·55) and breakfast skipping (adjusted mean, 95 % CI: Q1: 4·03, 3·81, 4·26; Q3: 3·43, 3·17, 3·69). Time in bed and sleep variability were associated with few eating behaviours.

Some, but not all, sleep indices were related to problematic eating behaviours. Sleep habits may be important to address in interventions and policies that target improvements in eating patterns and health outcomes.