OBJECTIVE: Breast-feeding may be more difficult in women with diabetes because of neonatal morbidity and fluctuating maternal blood glucose values. The frequency of long-term breast-feeding and the possible predictors for successful breast-feeding were investigated. RESEARCH DESIGN AND METHODS: One hundred two consecutive women with type 1 diabetes were interviewed about breast-feeding using a semistructured questionnaire 5 days and 4 months after delivery. Clinical data were collected from the medical records. RESULTS: Five days after delivery, 86% of the women were breast-feeding. Four months after delivery, 55 (54%) women were exclusively, 14 (14%) were partly, and 33 (32%) were not breast-feeding compared with 50, 26, and 24% in the background population (NS). Mothers exclusively breast-feeding at 4 months were characterized by previous experience with breast-feeding, a higher educational level, and vaginal delivery and included a high proportion of nonsmokers, whereas there were no associations with diabetes-related parameters such as white classes, duration of diabetes, HbA(1c), and insulin dose at conception. Breast-fed offspring had a significantly higher birth weight and gestational age and were less often receiving glucose intravenously compared with the remaining offspring. Independent predictors of exclusive breast-feeding at 4 months were previous experience with breast-feeding (odds ratio 6.3 [95% CI 2.4-17]) and higher educational level (7.1 [2.4-21]). Cessation of breast-feeding was mainly due to common nursing problems, such as perceived milk supply, and not related to maternal diabetes status. CONCLUSIONS: The majority of the women with type 1 diabetes initiated breast-feeding, and the prevalence of breast-feeding at 4 months was comparable to that in the background population. Independent predictors of exclusive breast-feeding at 4 months were previous experience with breast-feeding and higher educational level.