In mice, retrovirus-associated breast cancers are promoted by immune mechanisms, and immunosuppression during the premalignant phase reduces the incidence of breast cancer and prolongs life. If some women likewise have immune promotion of breast cancer, the incidence of breast cancer in patients receiving therapeutic immunosuppression should be lower than that in a comparable cohort of non-immunosuppressed women. We examined the incidence of de-novo breast cancer arising in women receiving immunosuppressive therapy after kidney or heart transplantation, comparing the figures with published rates. In 25,914 immunosuppressed women followed for 1-11 years there were 86 cases of breast cancer compared with 113.8 expected (p = 0.009). Incidence was particularly low in the first transplant year with relative risk 0.49, rising to 0.84 in subsequent years. For all other major cancers the incidence was higher in the immunosuppressed women. If, as in mice, the reduced incidence of breast cancer is a direct effect of immunosuppression, these observations raise the possibility of therapeutic manipulation of specific immune mechanisms that promote tumour growth.