One hundred and fifty-four parous women with pathologically confirmed newly diagnosed breast cancer in National Taiwan University Hospital were selected as the case group. Three hundred and eighty-six parous inpatient controls were individually matched for each case by age and date of admission. Information was collected from questionnaire interview and medical record. Effects of lactation and lactation suppression hormone were assessed with conditional logistic regression analysis. In univariate analysis, lactation had a weak protective effect, after adjusting for educational level, body mass index, family history of breast cancer, reproductive factors, oral contraceptive and lactation suppression hormone use, the weak negative associations changed to weak positive ones. The use of lactation suppression hormone was consistently associated with a higher breast cancer risk. The major confounders between lactation and breast cancer risk were parity and age at first full-term birth. Lower breast cancer risks were found in high parity - long lactation, and in younger ages at first full-term birth - long lactation groups. The differences among different lactation lengths within each parity and age at first full-term birth category were small. The protective effects disappeared after adjusting for other potential confounders. The results of this study questions the independent protective effect of lactation in a population having both a variable duration of lactation and rapid socioeconomic change. The role of lactation suppression hormones needs further study.