BACKGROUND: The evidence on the association between sedentary behaviour and depression in adolescence is mixed. We aimed to investigate the association between mentally-active and mentally-passive sedentary behaviours at 11 years (11y) and depressive symptoms at 14y, and to examine potential mediators. METHODS: UK Millennium Cohort Study data were used (n=7,124; 49% boys). At 11y, participants self-reported frequency of mentally-passive (listening to music, internet use) and mentally-active sedentary behaviours (reading, playing electronic games). Additional parental-reported behaviours (mentally-passive: TV viewing; mentally-active: homework) were summed with self-reported behaviours to represent continuous indicators of mentally-active and mentally-passive sedentary behaviour. Depressive symptoms were assessed (at 11y and 14y) using the short-version of Mood and Feelings Questionnaire. Body mass index (BMI), mentally-passive sedentary behaviour and cognition at 14y were examined as potential mediators. Linear regression models were adjusted for confounders and stratified by sex. Subsequent mediation analyses reporting e-values were used to assess unmeasured confounding. RESULTS: Among girls, mentally-passive sedentary behaviour at 11y was associated with later depressive symptoms (14y) [β:0.089 (95%CI:0.055-0.122), e-value:1.32]. This association was mediated by BMI [5.6% (95%CI:4.1%-8.6%)] and mentally-passive sedentary behaviour [105.6% (95%CI:79.6%-156.7%)]. No associations were observed in boys or between mentally-active sedentary behaviour and later depressive symptoms. LIMITATIONS: The parental report of behaviours and the assessment of mediators and outcome in the same wave are the main limitations. CONCLUSION: Future interventions aiming to improve mental health among girls could aim to reduce mentally-passive sedentary behaviour in early teens and could target potential mediators including BMI.