Background: The benefits of breastfeeding for infants and mothers have been well established, yet rates of breastfeeding remain well below national recommendations in the United States and even lower for women who smoke during pregnancy. Primary goals of this study were to explore contextual factors that contribute to breastfeeding intentions and behavior and to examine how smoking status affected women's decision making about breastfeeding. Methods: This paper is based on a longitudinal qualitative study of smoking, pregnancy, and breastfeeding among 44 low-income women in the southwest U.S. who smoked during pregnancy. Each woman was interviewed 9 times; 6 times during pregnancy and 3 times postpartum using semistructured questionnaires. Interviews lasted 1 to 3 hours and were tape-recorded, transcribed, and analyzed. Results: Despite 36 (82%) respondents stating that they intended to breastfeed for an average duration of 8 months, rates of breastfeeding initiation and duration were much lower than intentions. By 6 months postpartum, only two women were breastfeeding exclusively. Conclusions: Women perceived that a strong risk of harming the baby was posed by smoking while breastfeeding and received little encouragement to continue breastfeeding despite an inability to stop smoking. The perceptions of the toxic, addictive, and harmful effects of smoking on breastmilk constitution and quantity factored into reasons why women weaned their infants from breastfeeding much earlier than the recommended 6 months. The results indicate a need for more consistency and routine in educating women on the relationship between smoking and breastfeeding and in promoting breastfeeding in spite of smoking postpartum.