PURPOSE: Young caregivers experience, on average, poorer mental health outcomes than non-caregiving young people. However, it is unknown to what extent these effects differ with age, or among short-term versus long-term caregivers. Using repeated assessment of young caregiving across multiple waves of a prospective adolescent cohort study, we conducted repeated cross-sectional analyses of caregiver status and contemporaneous depressive symptoms, self-harm, and suicidality in early to middle adolescence. METHODS: Four waves of questionnaire data from a large, longitudinal population-based cohort study (Tokyo Teen Cohort) were analyzed. Caregiver status was collected from participants aged 10, 12, 14, and 16 years. Mental health outcomes assessed were depressive symptoms, self-harm and suicidal feelings. Logistic regression analyses were conducted assessing effects of a) young caregiver status and b) new, long-term, and ex-caregiver 2-year categorizations on mental health outcomes at 12, 14, and 16 years, both unadjusted and adjusted for potential confounders (sex, low income, single-parent household, and parental distress). RESULTS: Depressive symptoms were elevated among long-term caregivers at 14 years (unadjusted odds ratio (uOR): 3.11 [1.33-7.27], adjusted odds ratio (aOR): 2.49 [1.03-5.99]). Borderline associations between long-term caregiving and self-harm (uOR: 3.14 [1.06-9.35], aOR: 2.51 [0.82-7.63]) and suicidal feelings (uOR: 2.49 [0.98-6.34], aOR: 2.06 [0.80-5.33]) were detected at 16 years. No associations were found at age 12 years in primary analyses; sensitivity analyses indicated possible increased depressive symptoms. DISCUSSION: Young caregivers with long-term caregiving roles are at the greatest risk for negative mental health outcomes, with effects concentrated in later adolescence. These findings highlight urgent need for early identification and practical and psychological support for young people shouldering caregiving burdens.