AbstractObjective:To investigate the relationship between ultra-processed food consumption and early childhood caries.Design:Cross-sectional analysis of baseline data from a cluster randomised controlled study. Outcomes included the prevalence of children with non-cavitated and cavitated caries. The main exposure was the total daily consumption of ultra-processed foods (up to three times and four times or more), assessed through a FFQ. Potential confounders were socio-demographic characteristics of the child and caregiver/family, child breast-feeding, oral hygiene and use of dental services. Poisson regression using robust variance adjustment was used to estimate prevalence ratios (PR) and their respective 95 % CI.Setting:Primary Healthcare Centers in an urban area of Pelotas, Southern Brazil.Participants:Children aged 0–3 years (n 309).Results:Consumption of ultra-processed foods four times or more a day was found in 67·6 % of children; 24·4 and 12·0 % presented non-cavitated and cavitated caries, respectively. After adjustment, children who consumed ultra-processed foods four times or more a day were more likely to present both non-cavitated caries (PR 2·25, 95 % CI 1·19, 4·27, P = 0·013) and cavitated caries (PR 3·48, 95 % CI 1·18, 10·30, P = 0·024) compared with those who have consumed them up to three times a day.Conclusions:Consumption of ultra-processed foods is associated with early childhood caries. Interventions aiming at reducing ultra-processed food consumption should be implemented to improve children’s oral health.