THE possible existence of a relationship between thyroid activity and both the incidence and progress of breast cancer has been the subject of considerable discussion over a number of years. We reviewed the literature in our paper of 1961, and described the early stages of an experimental approach we were making in an attempt to answer the question (Sicher and Waterhouse, 1961 More than one type of relationship between thyroid function and breast cancer has been considered in the recent literature, and it is not always clear what is under investigation in a particular study. The hypothesis that hyperfunction of the thyroid gland is associated with a reduced incidence of breast cancer is supported by Humphrey and Swerdlow (1964) who found no case of breast cancer among 196 patients with hyperthyroidism followed for 12 years. Stoll (1962), however, cannot support the hypothesis from his study of 150 cases of breast cancer. Capelli and Margottini (1964) failed to detect a decrease in thyroid function among patients with breast cancer. Humphrey and Swerdlow (1964) found, among cases of breast cancer having a history of hyperthyroidism, both a higher 5-year survival rate and a lower incidence of local recurrences than among the remainder of their cases of the disease.The last finding mentioned above seems in conflict with that of Edelstyn, Lyons and Welbourn (1958) that breast cancer patients with only local extensions of growth had consistently higher indices of thyroid activity than patients in whom blood-borne spread had occurred. Reeve et al. (1961) reanalysed the data of Edelstyn et al. (1958) making certain plausible assumptions, and arrived at the same conclusions, although in a study of their own, using groups of patients similar to those of Edelstyn et al., they could show no difference in levels of thyroid activity between patients with local disease or with blood-borne metastases.Similar hypotheses have formed the basis of treatment in cases of breast cancer. Stoll (1962) in a series of 12 advanced cases treated with a combination of oestrogen and tri-iodothyronine (T3) could find no evidence for any regression ascribable specifically to T3. Emery and Trotter (1963) used T3 in a controlled study of 54 advanced cases without finding evidence of any noticeable effect on the prognosis. Lyons and Edelstyn (1965) used desiccated thyroid extract and later thyroxine in comparison with a control series and again could find no evidence of prognostic value in the treatment. They did find however an increased incidence