Abstract Background: There is little knowledge of the association between the structural setting of the school and irregular lunch consumption among adolescents. Objectives: To study whether the structural setting of the school was associated with adolescents’ irregular lunch consumption in a multilevel model controlled for factors at the individual level. Further, we investigated whether grade modified the association. Methods: We used data from the Danish 2010 contribution to the international collaborative cross-sectional study ‘Health Behaviour in School-Aged Children (HBSC). Data collection were conducted in 2010 among schoolchildren aged 11, 13 and 15 years (in Denmark, equivalent to 5th, 7th and 9th grade, respectively) from a random sample of schools, i.e. cluster sampling. The schoolchildren completed the self-administered and internationally standardised anonymous HBSC questionnaire at school (response rate=86.3%) and the school principal completed the principal questionnaire (response rate=94.5%). The study sample for this study comprised 4489 students within 275 classes within 69 schools. Multilevel logistic regression analyses were used to estimate the associations between school level variables and irregular lunch consumption. Lunch consumption was measured by lunch frequency questions. Results: We found that the school level factors, “Availability to canteen” and “Adult present in lunch breaks” were associated with irregular lunch consumption. Students with no access to canteen had a lower risk of having irregular lunch consumption OR=0.77 (0.62-0.96). Students with no adult present in lunch breaks had a higher risk of irregular lunch consumption OR=1.62 (1.32-1.99). Having irregular lunch consumption was associated with the following individual characteristics: boy, 7th grade, medium and low family social class, descendant, and single and reconstructed family type. In the final model 0.1% of the total variance was explained by differences between schools and 1.8% was explained by differences between classes in the same school. Grade did to some extent modify the association. Conclusion: The structural setting of the school was associated with children and adolescents’ regularity of lunch consumption in a multilevel model controlled for individual level factors. From the findings we can conclude that availability of canteen and school stall did not promote regular lunch consumption while the presence of an adult in lunch breaks promoted regular lunch consumption. Grade did to some extent modify the association. Most of the total variance in students’ lunch consumption was explained by individual level (98.2%).