Background: Vitamin D deficiency in pregnancy may be a risk factor for child allergic disease. However, less is known about disease risk across different levels of vitamin D. Objective: We aimed to examine the relation between a maternal vitamin D prediction score and child allergic disease. Methods: A total of 32,456 pregnant women were enrolled in the Danish National Birth Cohort (1996-2003) and had data on a validated vitamin D prediction score based on 1497 mid-pregnancy plasma 25(OH)D samples. Child allergic disease was assessed at 18 months and at 7 years using questionnaire data and national registry extracts. We used multivariable log-binomial models to quantify risk ratios (RR) and 95% CI. Plasma 25(OH)D was examined in a stability analysis. Results: Median (IQR) vitamin D prediction score was 58.7 (49.2-69.0) nmol/l. In main analysis, there was no association between vitamin D prediction score examined in quintiles or by restricted categories (≥75 nmol/l and <25 nmol/l vs. 25-74.9 nmol/l) and child allergic disease. However, maternal vitamin D prediction score ≥100 nmol/l(vs. 50-79.9 nmol/l) was associated with increased risks of child asthma at 18 months (RR: 1.36, 95% CI: 1.02, 1.80) and asthma by hospital admission (RR: 1.65, 95% CI: 1.04, 2.62). For vitamin D prediction score <25-30 nmol/l, there were increased risks of child asthma at 18 months and by hospital admission and medication prescription at age 7, although these findings were not robust to covariate adjustment. Similar results were found for plasma 25(OH)D. Conclusions: Our study provided little evidence for an association between maternal vitamin D prediction score and child allergic disease for scores ≥75 nmol/l. However, increased risks were observed for vitamin D prediction score ≥100 nmol/l. These associations are hypothesis generating and would need to be replicated in other cohorts.