Prenatal exposure to elements may be associated with birth weight via shortening of gestation. This study aimed to determine if prenatal exposure is associated with birth weight, and to explore the potential mediating role of gestational age in the association. Within an established Bangladesh prospective birth cohort (2008–2011), we analyzed the concentrations of 15 elements in maternal serum samples collected during the first (n = 780) and second (n = 610) trimesters using inductively coupled plasma mass spectrometry. Mediation analyses explored the relationships between these elements, gestational age, and birth weight. Serum concentrations of cobalt (Co) (first trimester: b = 56.5; 95% confidence interval [CI]: 13.5–99.0; false discovery rate [FDR]-q = 0.035; second trimester: b = 73.3; 95% CI: 20.4–130.2; FDR-q = 0.037) and antimony (Sb) in both trimesters (first trimester:b = 92.1; 95% CI: 66.0–118.9; FDR-q < 0.001; second trimester: b = 93.3; 95% CI: 67.3–118.4; FDR-q < 0.001), and strontium (Sr) in the first trimester (b = 142.4; 95% CI: 41.6–247.9; FDR-q = 0.035) were positively associated with birth weight, while negative associations were observed for barium (Ba) (first trimester: b = −154.8; 95% CI: −217.9 to 91.8; FDR-q <0.001; second trimester: b = −26.7; 95% CI: −44.9 to 10.2; FDR-q < 0.001). These elements act partially by affecting gestation age and appear to have heightened impact among smaller infants. Further research is needed to determine the biological underpinnings of these effects, which may inform strategies to avert low birth weight.