Introduction Diet diversity (DD) in infancy may be protective for early food allergy (FA) but there is limited knowledge about how DD incorporating consumption frequency influences FA risk. Methods Three measures of DD were investigated in 2060 infants at 6 and/or 9 months of age within the NorthPop Birth Cohort Study; a weighted DD score based on intake frequency, the number of introduced foods, and the number of introduced allergenic foods. In multivariable logistic regression models based on directed acyclic graphs, associations to parentally reported physician diagnosed FA at 9- and 18-months age were estimated, including sensitivity- and stratified analyses. Results High weighted DD scores (24-31p) at 9 months of age was associated with a 61% decreased FA risk at age 18 months [OR (95% CI) = 0.39 0.18-0.88] compared with infants with the lowest DD scores (0-17p). The association remained significant after exclusion of early FA cases. Having introduced 13-14 foods at age 9 months, independent of consumption frequency, was associated to a 45% decreased FA risk [OR (95% CI) = 0.55 (0.31-0.98)] compared to having introduced 0-10 foods. When stratifying, results remained only for children with no FA history in the family. No association was seen between reduced FA risk and DD measured at 6 months of age or having introduced more allergenic foods in infancy. Conclusion A diverse diet at 9 months of age may prevent FA at 18 months and results underscore the need for additional investigations on the impact of consumption frequency in infancy.