Background: It is necessary to determine whether there are adverse health effects of prenatal exposure to long-term, low levels of mercury and selenium. However, there are limited that reports on the association between mercury levels by selenium levels and birth size. Therefore, we examined whether maternal mercury levels during pregnancy had any effect on infant birth size, and size, and whether selenium levels influenced this relationship.Objectives: To examine the association between mercury and selenium levels during pregnancy with infant birth size.Methods: The Japan Environment and Children's Study is a prospective birth cohort conducted between 2011 and 2014. Total mercury levels and total selenium levels in maternal blood during the second and third tri-mesters were measured using Inductively Coupled Plasma-Mass Spectrometry. Birth weight and small-for-gestational-age were confirmed by medical records. Small-for-gestational-age was defined as birth weight below the 10th percentile according to standard percentile for gender, parity, and gestational age. Multiple linear and logistic regression analyses were used to examine the association between maternal mercury exposure and birth weight or small-for-gestational-age adjusted for confounders (including maternal age and body mass index pregnancy).Results: Overall, 15,444 pregnant women were included in this study. Median (inter-quartile range) of blood mercury and selenium levels were 3.66 (2.59-5.18) ng/g and 170.0 (158.0-183.0) ng/g, respectively. Compared to infants of mothers with the highest blood selenium level, those of mothers with the lowest blood selenium level had neither a significant birth weight increase (9 g, 95% confidence interval: -6, 25) nor a significant odds ratio for small-for-gestational-age (0.903, 95% confidence interval: 0.748, 1.089). Compared to infants of mothers with the lowest blood mercury level, those of mothers with the highest blood mercury level had neither a significant birth weight reduction (-12 g, 95% confidence interval: -27, 4) nor a significant odds ratio for small-for-gestational-age (0.951, 95% confidence interval: 0.786, 1.150). Compared to infants of mothers with the lowest quartile of maternal blood mercury level, all infants of mothers with the highest quartile of maternal blood mercury level had a reduced birth head circumference of 0.073 cm (95% confidence interval: -0.134, -0.011).Conclusions: There was no association between maternal blood mercury levels and small-for-gestational-age and birth weight among 15,444 pregnant women. In a Japanese population, which has a relatively higher blood mercury level than reported in Western population, reduced birth size was not found to be associated with blood mercury levels, with the exception of birth head circumference.