Mental illnesses emerge in childhood, making early intervention important. However, antidepressant treatment rates remain low following a controversial FDA warning. We provide some of the first evidence of impacts of antidepressant treatment in childhood on objectively measured mental health indicators and economic outcomes over time, and the first attempt to investigate under- vs overtreatment. Treatment improves Math scores in high school, post-compulsory education and adult employment and earnings, reducing welfare dependence. It reduces suicidality and hospital visits. Low-SES children benefit more. Policy simulations in a marginal treatment effects framework suggest under-treatment, highlighting that expanding treatment can reduce inequality.