Objective: To experimentally compare the effects of eating or skipping breakfast on energy expenditure, activity levels and dietary habits. Design: A randomised cross-over trial, lasting 2 weeks. Participants were provided breakfast during one week and were required to fast until mid-day during the other week. Setting: University campus. Subjects: Forty-nine participants (twenty-six female and twenty-three male participants) were recruited. Food intake was monitored using food diaries, and energy expenditure was assessed using pedometers and heart rate monitors. Morningness-eveningness, physical activity and health were assessed using validated questionnaires. Results: Across all participants, daily energy expenditure did not differ between the two experimental conditions. Total energy intake over 24 h did not vary with condition (male participants: 8134 (SD 447) kJ/d and 7514 (SD 368) kJ/d; female participants: 7778 (SD 410) kJ/d and 7531 (SD 535) kJ/d, for the breakfast and no-breakfast conditions, respectively). However, when comparing habitual breakfast eaters with those with irregular or breakfast-skipping habits, it was found that male non-habitual breakfast eaters consumed significantly (P 5 0?029) more energy during the breakfast condition. Furthermore, female participants who were habitual breakfast eaters were found to eat significantly (P 5 0?005) more and later in the day under the no-breakfast condition. Conclusions: Although the suggestion that breakfast is a behavioural marker for appropriate dietary and physical activity patterns is not refuted by the present findings, our data suggest that the effect of breakfast may vary as a function of gender and morning eating habits, and thus there may be other mechanisms that link BMI and breakfast consumption behaviour.