Attention deficit hyperactivity disorder (ADHD), a prevalent neurodevelopmental disorder in children, originates from a multifaceted interplay of genetic, neurological, and environmental factors. Recent studies have increasingly concentrated on environmental determinants, notably air pollution, and their impact on the risk of developing ADHD. Additionally, previous research has often conflated clinically diagnosed ADHD cases with instances of mere ADHD-like symptoms, a methodology that can introduce bias and obscure the true relationship between environmental factors and ADHD. To address this oversight, our systematic review meticulously investigates the relationship between both prenatal and postnatal exposures to particular air pollutants and strictly clinically diagnosed ADHD. Our comprehensive review encompassed 801 studies from PubMed, Cochrane Library, Web of Science, and Embase databases, out of which eight met our rigorous inclusion criteria. The Newcastle-Ottawa Scale (NOS) was utilized to gauge quality and bias. Our review found substantiated the connection between prenatal exposure to PM2.5 and NOx and a heightened risk of ADHD, while exposure to PM10 during the prenatal stage was not associated with ADHD. These findings hint at varied health impacts from different particulate matters and the prospect of gender-specific susceptibilities to such exposures. We also identified an association between postnatal exposure to PM2.5, PM10, and NO2 and an increased ADHD risk, underlining the potential neurodevelopmental harms from early exposure to these pollutants. These relationships, seemingly intricate and potentially dose-dependent, underscore the need for more detailed scrutiny. The unique value of our review is in its detailed exploration of the association between specific air pollution exposures and clinically diagnosed ADHD. Our findings offer much-needed clarity in this complex domain and emphasize the importance of future research to standardize exposure and outcome metrics, probe potential mechanisms, and reduce bias and heterogeneity.