Some recent human and animal studies have suggested that air pollution may affect the central nervous system and contribute to neurodevelopmental outcomes, such as autism spectrum disorder (ASD). We aimed to investigate the association between long-term exposure to ambient air pollution and increased odds of ASD among 2 to 10-year-old children. We conducted a case-control study in Tehran, Iran. Cases were 134 children born between 2004 and 2012 diagnosed with ASD whose mothers were resident in Tehran during their pregnancy, and controls were 388 children without ASD randomly selected from public schools and kindergartens. Land-use regression models were used to estimate their annual mean exposure to ambient particulate matter with aerodynamic diameter ≤10 μm (PM10), sulfur dioxide (SO2), benzene, toluene, ethylbenzene, p-xylene, o-xylene, m-xylene (BTEX), and total BTEX. Logistic regression was used for the analyses and adjusted for possible confounding variables. The odds ratios per 1 unit increase in pollutants in the adjusted models were 1.00 (95% CI: 0.99, 1.01) for PM10, 0.99 (95% CI: 0.99, 1.00) for SO2, 0.96 (0.83, 1.11) for benzene, 1.00 (0.96, 1.04) for toluene, 0.95 (0.79, 1.16) for ethylbenzene, 1.00 (0.78, 1.27) for p-xylene, 1.09 (0.94, 1.27) for o-xylene, 1.01 (0.92, 1.12) for m-xylene, and 0.99 (0.97, 1.01) for total BTEX. We did not find the evidence of association between estimated annual mean exposure to abovementioned ambient air pollutants and increased odds of ASD in children. However, our findings might be due to some important limitations. Further research with better control of confounding variables, improved spatiotemporal exposure estimates, and inclusion of other important markers of air pollution is recommended.