Purpose: To explore predictors of breastfeeding behavior and seek to replicate prior findings using a recent nationally representative United States (US) sample, which has not been done previously.Design: Secondary data analysis using the National Survey of Children's Health 2011 to 2012.Setting: Households in the 50 US states, the District of Columbia, and the US Virgin Islands.Participants: Caregivers with a child between 6 months and 6 years of age (n = 26 552).Measures: Caregiver-reported child breastfeeding behavior, child birth weight, child/maternal/family demographics, maternal physical and mental health, and household tobacco exposure.Analysis: A multinomial logistic regression explored whether child, maternal, and family characteristics predicted exclusive breastfeeding for 6 months or breastfeeding for a suboptimal duration.Results: Approximately 79% and 17% of the sample-initiated breastfeeding and exclusively breastfed for 6 months, respectively. Child ethnicity/race, maternal education, household tobacco exposure, family composition, and family income predicted odds of both exclusive breastfeeding for 6 months and breastfeeding for a suboptimal duration or not exclusively. Normal or high child birth weight and better maternal physical health predicted greater odds of exclusive breastfeeding for 6 months.Conclusion: Rates of exclusive breastfeeding for 6 months remain low in the United States. Single-parent and step-families, lower income families, non-Hispanic black children, children with exposure to tobacco, and children of mothers with lower education are at greatest risk. [ABSTRACT FROM AUTHOR]