Between 1972 and mid-1990 the frequency of breast cancer was studied in a group of 261 mostly premenopausal women of the gynaecological department of the Municipal Hospital in The Hague, the Netherlands. All the patients had had a total hystero-adnexectomy after which they were substituted with estradiol implants. On the basis of a stratified lifetable giving the cumulative incidence of breast cancer in the Netherlands, an expected incidence of 2 per 1000 person-years was estimated for the observed group (mean observation period: 8.25 years). There were three cases of breast cancer in the observed group. This means an incidence density of 1.4 per 1000 person-years. It is concluded that this form of oestrogen substitution does not increase the risk of breast cancer.