The risk of breast cancer associated with delivering a twin birth was examined in a population-based nested case-control study of nearly 4800 Swedish women with breast cancer and 47000 age-matched control subjects. All were aged less than 50 years and parous. After adjustment for age at first birth and parity, a 29% reduction in breast cancer risk was observed in mothers of twins relative to those who were not (odds ratio = 0.71, 95% confidence interval 0.55-0.91). These results provide evidence that women who bear twins are at reduced risk of breast cancer, one explanation for which may be their unusual levels of hormonal exposure.