Although introduction of inappropriate foods and liquids in early childhood and their association with breastfeeding is commonly reported in US children, coffee use in very young US children and its association with breastfeeding is not. This study aimed to determine the proportion of 1- and 2-year-olds in an urban population consuming coffee, their rate of consumption, and predictors of consumption, including breastfeeding status. We used data from a prospective cohort study on infant weight gain and diet, and body mass index at age 2. We used bivariate analyses to examine variables associated with coffee consumption at 1 and 2 years and multivariate logistic regression to control for variables of interest. This study included 315 mother-infant dyads. At 1 year, the rate of coffee consumption reported was 2.5%; at 2 years, it was 15.2% and average daily consumption was 1.09 oz (range, 0.01- 4.00 oz). The only characteristic associated with coffee consumption at 1 year was breastfeeding at 1 year (P = .0275), which did not remain significant after controlling for confounding variables. Variables significantly associated with coffee consumption at year 2 were lower maternal education (P = .0016), non-US maternal place of birth (P = .0015), maternal Hispanic ethnicity (P < .0001), infant female sex (P = .0495), and receiving any breast milk at 1 year of age (P = .0189). After multivariate logistic regression, maternal Hispanic ethnicity (P = .0139) and infant female sex (P = .0371) remained significant. Coffee consumption is not uncommon among toddlers in Boston, Massachusetts, USA. After controlling for possible confounding factors, maternal ethnicity and infant sex were significantly associated with this practice.