Background. We aimed to evaluate the association between maternal smoking and the occurrence of childhood refractive error and amblyopia. Methods. Relevant articles were identified from PubMed and EMBASE up toMay 2015. Combined odds ratio (OR) corresponding with its 95% confidence interval (CI) was calculated to evaluate the influence of maternal smoking on childhood refractive error and amblyopia. The heterogeneity was evaluated with the Chi-square-based Q statistic and the.. 2 test. Potential publication bias was finally examined by Egger's test. Results. A total of 9 articles were included in this meta-analysis. The pooled OR showed that there was no significant association between maternal smoking and childhood refractive error. However, children whose mother smoked during pregnancy were 1.47 (95% CI: 1.12-1.93) times and 1.43 (95% CI: 1.23-1.66) times more likely to suffer from amblyopia and hyperopia, respectively, compared with children whose mother did not smoke, and the difference was significant. Significant heterogeneity was only found among studies involving the influence of maternal smoking on children's refractive error (P < 0.05;I-2 = 69.9%). No potential publication bias was detected by Egger's test. Conclusion. The meta-analysis suggests that maternal smoking is a risk factor for childhood hyperopia and amblyopia.