BackgroundMethods to objectively assess fruit and vegetable (FV) intake in young children are needed in order to support rigorous assessments of policies and interventions. Noninvasive skin carotenoid concentration measurements may provide a rapid assessment of toddler carotenoid and carotenoid-rich FV intake. ObjectiveThe aim of this study was to evaluate the relationship between pressure-mediated reflection spectroscopy-measured skin carotenoid score (SCS) with proxy-reported carotenoid and FV intake in racially and ethnically diverse US toddlers. DesignThis study was a secondary analysis of data obtained from a randomized controlled 10-week study of the effect of an interactive healthy playgroup intervention vs classroom parent education on the diet and physical activity of toddlers. This study collected skin carotenoid and 1-week dietary intake using a 31-item semi-quantitative food frequency questionnaire analyzed by the Nutrient Data System for Research nutrient database. The current study determined dietary predictors of toddler SCS. Participants/settingParticipants were racially and ethnically diverse US toddlers (12 to 36 months) and adult guardian dyads (n = 50) recruited from the community in Houston, TX from fall 2018 to spring 2019. Main outcome measuresAt baseline and 10 to 12 weeks after baseline, SCS was measured by means of pressure-mediated reflection spectroscopy (Veggie Meter, Longevity Link Corporation), guardians reported toddler diet using a food frequency questionnaire, and carotenoid intake was estimated from the food frequency questionnaire responses using a nutrient database. Statistical analysisThe relationship between toddler SCS and intake of total and individual carotenoid species and FV servings was tested using generalized linear mixed models, controlling for body mass index-for-age percentiles, group assignment, and age. ResultsSCS was positively and significantly predicted by estimated intakes of total carotenoids (P = .002), β-carotene (P ≤ .001), and lutein and zeaxanthin (P = .003). Reported intakes of α-carotene, β-cryptoxanthin, and lycopene were not predictors of SCS. SCS was predicted by estimated total FV intake (P = .047) and vegetable intake (P = .006), but not fruit intake (P = .580). ConclusionsThese results showed that reported dietary carotenoid intake is a significant predictor of SCS in an ethnically and racially diverse population of US toddlers. Toddler skin carotenoid measurement holds promise as a rapid, objective, noninvasive biomarker of dietary carotenoid intake.