Background: Isoflavone (ISO) consumption during pregnancy may affect food allergy (FA) in offspring through estrogen‐like and other biological effects; however, this ISO–FA association in mother–child pairs remains elusive. Herein, the maternal ISO–childhood FA association was investigated using the Japanese birth cohort data. Methods: The data from the Japan Environment and Children's Study, including 86,012 and 75,307 mother–child pairs, with FA in children aged ≤1 year and 1–3 years, respectively, were analyzed. Maternal ISO consumption (total daidzein and genistein) was assessed using a food frequency questionnaire. Physician‐diagnosed FA data were obtained using questionnaires completed by caregivers. ISO consumption data was categorized into quartiles (Q1: reference), and the ISO–FA association was assessed through multivariable logistic regression analysis after adjusting for maternal age, educational level, smoking status, pre‐pregnancy body mass index, history of allergies, blood folic acid concentration, total energy intake, and infant feeding method up to 4 months of age. Results: Overall, FA cases in children aged ≤1 and 1–3 years were 6.6% and 12.1%, respectively. Maternal ISO consumption was negatively associated with FA in all children aged ≤1 year (Q3, adjusted odds ratio [aOR]: 0.91, 95% confidence interval [CI]: 0.84–0.99; Q4, aOR: 0.86, 95% CI: 0.79–0.93) and in those aged 1–3 years (Q3, aOR: 0.93, 95% CI: 0.88–0.99; Q4, aOR: 0.88, 95% CI: 0.83–0.95). Furthermore, in sex‐stratified analysis, a negative association was observed between maternal ISO consumption and FA in female children of all ages, whereas this inverse association was observed only in 1–3‐year‐old male children. Conclusions: Overall, the findings show that maternal ISO consumption during pregnancy may reduce the risk of childhood FA, exhibiting sex‐difference effects on offspring development.