Lactation might have a crucial role in an extraordinary increase in breast cancer incidence in Korea, as the proportion of mothers who practised breast-feeding fell dramatically. This hospital-based case-control analysis has been carried out since 1997 to evaluate whether lactation is associated with breast cancer risk in Korean women. Among the eligible study participants, a total of 753 histologically confirmed incident cases and an equal number of controls were included in the analysis. The risk was estimated using unconditional logistic regression models. Family history, older at menopause, more full-term pregnancies increased the risk of breast cancer. Breast cancer risk decreased according to the total months of breast-feeding (P for trend=0.03). Average duration of breast-feeding of 11-12 months reduced risk of breast cancer by 54% compared with the duration of 1-4 months (odds ratio, 0.46; 95% confidence interval, 0.30-0.70). The decreasing risk trend according to average months of breast-feeding was also statistically significant (P for trend=0.02). Moreover, a reduced risk of breast cancer was apparent when analysis was restricted to the first breast-fed child (P for trend=0.006). This study confirms that lactation has an apparent dose-dependent protective effect against breast cancer in Korean women.