Objective To investigate associations between exposure to air pollution and child and adolescent mental health.

Design Observational study.

Setting Swedish National Register data on dispensed medications for a broad range of psychiatric disorders, including sedative medications, sleeping pills and antipsychotic medications, together with socioeconomic and demographic data and a national land use regression model for air pollution concentrations for NO₂, PM₁₀ and PM_2.5.

Participants The entire population under 18 years of age in 4 major counties. We excluded cohort members whose parents had dispensed a medication in the same medication group since the start date of the register. The cohort size was 552 221.

Main outcome measures Cox proportional hazards models to estimate HRs and their 95% CIs for the outcomes, adjusted for individual-level and group-level characteristics.

Results The average length of follow-up was 3.5 years, with an average number of events per 1000 cohort members of ∼21. The mean annual level of NO₂ was 9.8 µg/m³. Children and adolescents living in areas with higher air pollution concentrations were more likely to have a dispensed medication for a psychiatric disorder during follow-up (HR=1.09, 95% CI 1.06 to 1.12, associated with a 10 µg/m³ increase in NO₂). The association with NO₂ was clearly present in 3 out of 4 counties in the study area; however, no statistically significant heterogeneity was detected.

Conclusion There may be a link between exposure to air pollution and dispensed medications for certain psychiatric disorders in children and adolescents even at the relatively low levels of air pollution in the study regions. The findings should be corroborated by others.