Part A: Tamoxifen is widely used as first-time endocrine therapy in the treatment of breast cancer and women may be exposed to many years of treatment. The aim of this part of the study was to add to the body of evidence available on the risk to benefit profile of long-term tamoxifen use. The method was linkage of the Scottish adjuvant tamoxifen trial database with Scottish hospital inpatient statistics (SMR1), cancer registration (SMR6) and death records from the General Registry Office (GRO). The Scottish adjuvant tamoxifen adjuvant tamoxifen trial recruited 1323 patients between 1978 and 1984 and includes women who have received up to 18 years of tamoxifen therapy. The data linkage enabled ascertainment of causes of admission to hospital, registration of new primary cancers and causes of death in the trial population to the end of March 1996. Analysis by Cox Proportional Hazards method considered tamoxifen status as 'never use', 'ever use' or 'current use'. The results confirm a reduced incidence of ischaemic heart disease and increased thromboembolism. An increased risk of cataract in tamoxifen users was also detected. The effect of long-term tamoxifen therapy on the incidence of new primary malignancies and fractures is also discussed. Part B: The Edinburgh Breast Unit database has a record for every woman presenting to the department. The details recorded include previous exposure to exogenous oestrogens in the form of hormone replacement therapy (HRT) and oral contraception (OC). Many prospective epidemiological studies have looked at the effects of long-term exposure to exogenous oestrogens, mostly using annual or biennial questionnaires to determine events, with the inevitable problems of incomplete ascertainment. Data linkage in Scotland provides an alternative means of ascertaining events and offers an opportunity to study the effects of oestrogen exposure on a range of pre-defined endpoints.