Background Recent studies showed that air pollution might play a role in the etiology of mental disorders. In this study we evaluated the association between air pollution and mental and self-rated health and the possible mediating effect of physical activity in this association. Methods In 2008, 2013 and 2018 the Belgian Health Interview Survey (BHIS) enrolled 16,455 participants who completed following mental health dimensions: psychological distress, suboptimal vitality, suicidal ideation, and depressive and generalized anxiety disorder and self-rated health. Annual exposure to nitrogen dioxide (NO2), particulate matter <= 2.5 mu m (PM2.5) and black carbon (BC) were estimated at the participants' residence by a high resolution spatiotemporal model. Multivariate logistic regressions were carried out taking into account a priori selected covariates. Results Long-term exposure to PM2.5, BC and NO2 averaged 14.5, 1.4, and 21.8 mu g/m(3), respectively. An interquartile range (IQR) increment in PM2.5 exposure was associated with higher odds of suboptimal vitality (OR = 1.27; 95% CI: 1.13, 1.42), poor self-rated health (OR = 1.20; 95% CI: 1.09, 1.32) and depressive disorder (OR = 1.19; 95% CI: 1.00, 1.41). Secondly, an association was found between BC exposure and higher odds of poor self-rated health and depressive and generalized anxiety disorder and between NO2 exposure and higher odds of psychological distress, suboptimal vitality and poor self-rated health. No association was found between long-term ambient air pollution and suicidal ideation or severe psychological distress. The mediation analysis suggested that between 15.2% (PM2.5-generalized anxiety disorder) and 40.1% (NO2-poor self-rated health) of the association may be mediated by a difference in physical activity. Conclusions Long-term exposure to PM2.5, BC or NO2 was adversely associated with multiple mental health dimensions and self-rated health and part of the association was mediated by physical activity. Our results suggest that policies ...