This study was aimed at assessing the effects of therapeutic long-term administration of androgens on normal human female breast. Sections from mastectomy specimens of 29 female-to-male transsexuals who had prolonged androgen administration prior to surgery were examined using routine light microscopy and immunohistochemical techniques. For comparison, sections from ten 'normal' female breast reduction mammoplasty specimens and ten cases of gynaecomastia were similarly examined. Haematoxylin and eosin-stained sections were assessed for the prevalence of elements of the normal breast and benign breast lesions. Immunoperoxidase techniques were performed to study the distribution of a variety of breast-associated antigens and receptors. The results were assessed semi-quantitatively. The prevalence of normal acini and ducts, fibrosis, cysts, and apocrine metaplasia in transsexual specimens was not statistically different from that seen in normal controls. However, transsexual specimens had a significantly higher prevalence of microcalcification than normals. The majority of transsexual specimens were positive for gross cystic disease fluid protein-15, lactoferrin, and progesterone and oestrogen receptors, and negative for B72.3 and pS2. These findings were not significantly different from those in normal controls. All ten gynaecomastia specimens were positive for oestrogen and progesterone receptors. The prevalence of oestrogen receptors was significantly higher than that seen in transsexuals and normal controls, but the prevalence of progesterone receptors was only significantly higher than that seen in transsexuals. It is concluded that long-term androgen administration does not appear to have any significant lasting effect on the normal human female breast, as demonstrated by a wide range of histological and immunohistological criteria.