Abstract Context Vitamin D (VD) deficiency in pregnancy and the neonatal period has impacts on childhood outcomes. Maternal VD sufficiency is crucial for sufficiency in the neonate, though the effect of early versus late pregnancy 25-hydroxy-vitamin D (25(OH)D) levels on neonatal levels is unknown. Furthermore, chemiluminescence immunoassays (CLIAs) are widely used, though their validity in measuring 25(OH)D specifically in cord blood specimens has not been established. Objective To assess the validity of a CLIA in the measurement of cord blood 25(OH)D and to evaluate maternal determinants of neonatal 25(OH)D, including early versus late pregnancy 25(OH)D levels. Design This is an ancillary analysis from the Vitamin D Antenatal Asthma Reduction Trial (VDAART), a randomized, double-blinded, placebo-controlled study. Participants and Intervention A total of 881 pregnant women at high risk of having offspring asthma were randomized to receive VD supplementation or placebo. Serum samples were collected from mothers in early and late pregnancy and from offspring cord blood at birth. 25(OH)D levels were assayed by CLIA in all maternal and offspring samples and by LC-MS/MS in all offspring samples and a subset of 200 maternal third trimester samples. Results Cord blood 25(OH)D levels were higher as measured by CLIA (mean 37.13 ng/mL [SD 18.30]) than by LC-MS/MS (mean 23.54 ng/mL [SD 11.99]), with a mean positive bias of 13.54 ng/mL (SD 12.92) by Bland-Altman analysis. This positive bias in measurement by CLIA was not observed in maternal samples. Third trimester 25(OH)D was a positive determinant of neonatal 25(OH)D levels. Conclusion Chemiluminescence immunoassays overestimate 25(OH)D levels in human cord blood samples, an effect not observed in maternal blood samples. The quantification of 25(OH)D by CLIA should therefore not be considered valid when assayed in cord blood samples. Third trimester, but not first trimester, maternal 25(OH)D is one of several determinants of neonatal 25(OH)D status.