To explore the predictive value of cord blood 25(OH)D3 [25(OH)D3] for infantile atopic dermatitis (AD), and to provide a reference for primary prevention of early infantile AD. The neonates born from July to September, 2015 were enrolled. The cord blood samples were collected at birth to measure the level of 25(OH)D3. Outpatient follow-up was conducted for all the infants at 6 weeks, 3 months, and 6 months after birth. A survey was performed to investigate the incidence of AD. A total of 67 neonates completed a 6-month follow-up. The incidence of AD was 34% (23/67), and 91% (21/23) of these cases occurred in the first month after birth. The 23 AD children had a significantly lower cord 25(OH)D3 level than those without AD (P<0.05). The children with a cord 25(OH)D3 level <30 nmol/L showed a significantly higher incidence of AD than those with a cord 25(OH)D3 level ≥30 nmol/L (P<0.05). The receiver operating characteristic (ROC) analysis showed that the area under the ROC curve of cord 25(OH)D3 in predicting AD was 0.648 (standard error: 0.075; 95%CI: 0.502-0.795). Its sensitivity, specificity, positive predictive value, and negative predictive value were 52.2%, 79.5%, 57.1%, and 76.1%, respectively. Logistic regression analysis showed that low cord 25(OH)D3 level, preference for seafood during pregnancy, atopic family history, and mixed feeding were risk factors for infantile AD (P<0.05). Cord 25(OH)D3 level is inversely associated with the risk of infantile AD, but it has a low diagnostic value for this disease.