To evaluate the association between source of drinking water and dental caries at age 5 and to test whether socioeconomic conditions act as confounding factors in such association. The study was carried out in a sub-sample of the 2004 Pelotas Birth Cohort, which comprised the application of a questionnaire and clinical oral examination. The exposure was source of drinking water, collected through a question to the primary caregiver. The outcome was dental caries in primary dentition, measured through several standardized indicators of the decayed, missing and filled teeth index (dmft), assessing past and present dental caries. Socioeconomic indicators (family income and maternal education) were identified as potential confounding factors. After descriptive analysis, the association between source of water and measures of dental caries was assessed by Regression models. 1,084 children were evaluated and had complete information in all variables (83.2 % of the targeted sample). Dental caries experience was observed in 48.7 % of the children, with a mean of 1.9 decayed, missing or filled teeth. Most children consumed water from public water supply (76.0 %), and a socioeconomic pattern was observed, with children from lower income more likely to drink water from public supply. In crude analysis, children who consumed bottled water had a lower risk of decayed teeth, lower experience of dental caries and less severe disease. No associations were observed after adjustments for socioeconomic conditions. Underlying socioeconomic inequalities explained the association between prevalence and severity of dental caries in primary dentition and source of drinking water. Drinking fluoridated tap water is as effective in dental caries prevention as bottled water with acceptable levels of fluoride, with the advantage of being accessible to all. Oral health prevention and treatment should be implemented as early in life as possible and should take into consideration the family's socioeconomic context.