BACKGROUNDExposure to environmental pollutants during foetal and childhood development has been associated with neurodevelopmental disorders; however, existing evidence remains fragmented and lacks comprehensive credibility assessments.METHODSPubMed, Embase, PsycINFO, and the Cochrane Database of Systematic Reviews were systematically searched from database inception through January 2025, supplemented by reference list searches. Eligible studies were meta-analyses of observational research examining the associations between environmental pollutant exposure and adverse neurodevelopmental outcomes, with quality assessment performed using the AMSTAR 2 tool. Summary effect estimates were re-analysed using random-effects models, accompanied by heterogeneity I² statistics, 95 % prediction intervals, and evaluations of small-study effects and excess significance bias. Evidence was graded according to established criteria as follows: convincing, highly suggestive, suggestive, weak, or not significant. Reporting adhered to PRISMA guidelines, and the study protocol was pre-registered with PROSPERO (CRD42022339292).RESULTSA total of 45 studies, comprising 256 unique meta-analyses, were included in the umbrella review. Of these, 88 meta-analyses demonstrated statistical significance; however, 62 % of these exhibited small-study effects and/or excess significance bias. According to the quantitative grading criteria, highly suggestive evidence was identified for the association between PM2.5 exposure and attention deficit hyperactivity disorder (ADHD) (odds ratio [OR] 1.82 [1.52, 2.18]), as well as for lead (Pb) exposure (OR 1.96 [1.57, 2.46]) and ADHD. Childhood fluoride exposure was also graded as highly suggestive evidence for cognitive deficit (OR 3.80 [2.42, 5.97]). Suggestive evidence included the associations between pregnancy O3 exposure and autism spectrum disorder (ASD), childhood lead exposure with decreased intelligence quotient (IQ), decreased blood magnesium (Mg) levels in children with ADHD, and first-trimester pregnancy exposure to PM2.5 and ASD. Sensitivity analyses indicated that results derived from the Restricted Likelihood Maximum (REML) estimator demonstrated greater robustness compared to the widely used DerSimonian and Laird (DL) estimator, reflecting a 13 % increase in significant associations.CONCLUSIONS AND RELEVANCENeurodevelopmental disorders associated with environmental pollutants from industrial and human activities pose a critical public health challenge. Future research necessitates cautious analytical strategies and enhanced methodological transparency to improve the credibility of findings.