ABSTRACT: BACKGROUND: Results concerning the association between Body Mass Index (BMI) and depression in adolescence are conflicting, some describing a linear association (increase in BMI with level of depression), some a U-shaped association (both underweight and obesity are associated with high levels of depression), and they mostly concern small samples. The purpose of this study was to describe the association between BMI and depression in a large representative sample of French adolescents. METHODS: The association between BMI and depression, measured on the Adolescent Depression Rating Scale (ADRS), was tested in a French national representative sample of 39542 adolescents aged 17. Self-report data is derived from the 2008 ESCAPAD study, an epidemiological study based on a questionnaire focused on health and drug consumption.We used spline function analysis to describe the association between BMI and depression. RESULTS: The association between BMI and depression is significant (p<0.001) and non-linear for both genders, with no effect of parental working and marital status. For boys, there is U-shaped association. For girls the shape of the association is complex and shows inverted convexity for high levels of BMI. The spline shows higher scores for depression among overweight girls than among obese girls. CONCLUSION: There is evidence for a gender difference in the association between BMI and depression in adolescents, supporting the need to study boys and girls separately. Overweight adolescent girls are more likely to be depressed than obese adolescent girls, giving support for "fat and jolly" hypothesis not only among older women but also among adolescent girls.