Ninety-two patients with D2 prostatic cancer were studied. Initial treatment was with either diethylstilboestrol (DES) or orchidectomy. Response to DES (5/63, 81%) was significantly higher than for orchidectomy (18/29, 62%; p less than 0.01). However, duration of response and duration of survival were not significantly different for the 2 forms of hormonal therapy. Fifty-seven patients were randomised to receive second-line treatment with either medroxyprogesterone acetate (MPA), oral chlorambucil or combination chemotherapy (adriamycin + cyclophosphamide + 5-fluoro-uracil). Response to second-line treatment was similar for all 3 regimens (46% overall response). Most of the responses were disease stabilisation and, although there was symptomatic benefit, response to second-line therapy did not significantly improve survival compared to the survival experience of the group as a whole. It is concluded that palliative second-line treatment for advanced prostatic cancer should consist of the least toxic form of treatment which in this study was second-line hormone administration (MPA).