Using data from the Early Childhood Longitudinal Study (Birth Cohort 2001), this study investigates how the amount of time off from work after childbirth is associated with breastfeeding behavior, and how this association differs by maternal relationship status. Specifically, I observe how a delayed return to work after childbirth is associated with breastfeeding among a nationally representative sample of single, cohabiting and married working mothers. Using logistic regression and predicted probabilities, I find that returning to work after 12 weeks is associated with higher odds of breastfeeding beyond 3 months for married mothers. The amount of leave from work is not significantly associated with the breastfeeding behaviors of single or cohabiting mothers. These findings raise serious concerns about the design of current maternal leave policies, barriers in the workplace that discourage breastfeeding, and the ability of unmarried women to utilize benefits from a delayed to return to work after childbirth.