The possible relationship between breastfeeding and risk of breast cancer was assessed in a multinational hospital based case-control study that was conducted from 1979 to 1986 in 10 countries, including some in which many women lactated for long periods of time. Data were collected by personal interviews of 2338 cases and 14900 controls. No statistically significant trends in relative risk estimates with duration of lactation were observed in all women combined, in women who were pre- or postmenopausal, in women of various ages at diagnosis, in women with different numbers of livebirths, or in women characterized by the presence or absence of other risk factors for breast cancer. No trends in risk were observed with mean number of months that a woman lactated per breastfed child, or with number of pregnancies followed by treatment with oral or injectable medication to suppress lactation. In premenopausal women, and in those with two or more livebirths, most risk estimates for women who lactated for various lengths of time >6 months, relative to the risk in women who breastfed for ≤3 months, were less than unity, but the 95% confidence intervals of all estimates lncluded one. Long-term lactation may reduce slightly the risk of breast cancer, but the evidence for this from the present and prior investigations is not strong.