HUMAN and rat mammary tumours metabolize steroid hormones and thus have the capacity to modify their local hormonal environment (Adams & Wong, 1968;Jones et al., 1970;King et al., 1964;Miller et al., 1974). In previous studies, we have investigated the metabolism of testosterone by a predominantly ovary-dependent DMBA-induced mammary tumour, and have shown it to be sensitive to levels of other hormones present in vitro (Miller, 1976a), or in vivo (Miller, 1976b, c;Buchan et al., 1976). We now report a study of the metabolism of androgens by ovary-independent tumours derived by transplantation. By using multiple tumours transplanted to a single rat, it has been possible to examine tumours derived from the same host animal before and after 2 endocrine manipulations.All animals were of an inbred strain of Sprague-Dawley rat, obtained from the Animal Diseases Research Association (ADRA), Moredun Institute, Edinburgh. Two lines of transplantable tumours were used: TG3 and TG5. These were histologically carcinomas, originally induced in female "ADRA" rats at 50 days of age by i.v. administration of 5 mg of DMBA. Both tumours were then serially transplanted by dorsal skin implantation of tumour fragments into neonatally-thymectomized hosts. At the time of study TG3 was at its 6th passage and TG5 was at its 7th passage.Three separate portions of a single donor tumour were implanted at different dorsal sites in each neonatally thymectomized host animal at 50 days of age. Tumours were measured twice weekly throughout the period of study. When the largest tumour reached 1 5 x 1 5 cm in size it was removed through a dorsal skin incision and the animal was immediately bilaterally oophorectomized. Ten days later, a further tumour was excised. The animals were then given daily injections of oestradiol-17f (1 jig in 0-2 ml corn oil) for a further 10 days.The remaining tumour was excised and the animals were killed, no injection of oestradiol being given on the day of death. Eight rats were so treated (4 with TG3 tumours and 4 with TG5 tumours). Control animals (2 with TG3 tumours and 2 with TG5 tumours) were treated similarly except that a sham oophorectomy was performed in place of an oophorectomy and the injection vehicle replaced oestradiol.After excision, portions of each tumour (0.5 g) were used for steroid-metabolism studies. Each was finally sliced at 0°C in Krebs-Ringer phosphate buffer pH 7-4 (5 ml). An NADPH-generating system and radioactive precursor(20(lCi of either 7a(3H) dehydroepiandrosterone (DHA) or 7a(3H) testosterone) was added. The incubation systems were shaken at 37°C in an atmosphere of 02 for 2 h (TG3 series) or 1 h (TG5 series on account of the high activity found in the tumour lines). The reaction was stopped by adding methanol to 80% v/v and the incubations stored at