SUMMARYIn postmenopausal women with breast carcinoma, plasma and urinary oestrogens remain detectable following surgical adrenalectomy or hypophysectomy. These residual oestrogens could result from absorption of exogenous steroids, from endogenous production, or from a combination of these two sources. To determine whether endogenous production contributes to this oestrogen pool, we administered a potent steroidogenesis inhibitor, aminoglutethimide (AG), to women with breast carcinoma following hypophysectomy or adrenalectomy. Plasma and urinary oestrogens were measured with radioimmunoassays developed to provide appropriate sensitivity. In five women treated after initial hypophysectomy (hypox), plasma oestrone fell from 66·28 pg/ml (hypox) to 9·1 ± 2·4 pg/ml (hypox and AG) and oestradiol decreased from 8·3 ± 1·8 pg/ml to 2·5 ± 0·69 pg/ml. Similar decrements in urine oestrone (U‐E1) and ostradiol (U‐E2) were observed (U‐E1hypox: 2·25 ± 0·71 μg/24 h and 0·071·0±015 μg/24 h hypox and AG; U‐E20·47 ± 0·12 μg/24 h hypox to 0·124 ± 0·015 hypox and AG,P<0·05 for all). Similar significant reductions in plasma oestrone and oestradiol were observed in four women treated with aminoglutethimide following surgical adrenalectomy. While the levels of urinary oestrogens also fell in these patients, the differences were not statistically significant. In response to the decrements in oestrogen levels induced by AG, 2/5 women in the post‐hypophysectomy group and 2/4 in the post‐adrenalectomy group experienced partial objective tumour regression. These observations indicated that the residual oestrogens produced after surgical adrenalectomy or hypophysectomy, even though made in small quantities, were nonetheless biologically active. We conclude that endogenous production of oestrogens in extragonadal and extra‐adrenal sit