Continuous administration of gonadotrophin-releasing hormone (GnRH-)analogues leads to a receptor-down regulation of pituitary GnRH-receptors and subsequently inhibits ovarian hormone production. Since October, 1984, 118 evaluable pre- and perimenopausal patients (median age 42, range 25–55 years) with metastatic breast cancer were entered into an open phase II multicenter trial to evaluate efficacy of this new treatment modality. Patients were treated with the GnRH-analogue Goserelin (3.6 mg depot s.c. every 4 weeks) as first line therapy and followed up until progression. Mean serum gonadotrophins LH and FSH were significantly suppressed by Goserelin. Within 2–3 weeks, mean serum E2 values decreased to values seen in castrated women ( < 30pg/ml). Overall objective response with complete and partial remissions (CR + PR) was achieved in 44.9 % of patients with a median time to progression (mTTP) of 59 weeks, (range 20–163 weeks), no change (NC) in 28.0% with a mTPP of 27 weeks (range 16–101 weeks), and progression (P) in 27.1%. Responses were seen in ER-positive as well as ER-negative tumors, and in patients with different sites of metastases (loco-regional, bone, visceral, multiple). The value of different prognostic factors in relation to response rates, time to progression and time to death (overall survival) is discussed. Median overall survival (time from beginning of palliative Goserelin treatment to death) was 148 weeks. ‘Second line hormonal treatment’ of 41 patients (P after response to Goserelin first line therapy) with Tamoxifen (30 mg/d) in addition to ongoing Goserelin therapy showed CR + PR in 29 % of patients with a mTTP of 51 weeks (range 24–84 weeks), NC in 32% with a mTTP of 33 weeks (range 18–84 weeks), and P in 39%. ‘Third line hormonal treatment’ of 11 patients with additional amino-glutethimide to ongoing Goserelin therapy showed NC in 8 (73%) patients with a mTTP of 39 weeks (range 15–60 weeks), and P in 27%. Combination of Goserelin with chemotherapy during second or third line treatment showed higher response rates with a longer time to progression compared to cytotoxic treatment alone. All patients tolerated Goserelin monotherapy and combination treatments well without serious toxicity. In conclusion, Goserelin treatment of patients with metastatic breast cancer achieved response rates and duration of remissions which are at least comparable to those following oophorectomy. In premenopausal women, therefore, surgical castration as an irreversible procedure with psychological trauma and operative morbidity should be replaced by the treatment with GnRH-analogue depot preparations. Further randomised studies will clarify whether Goserelin monotherapy or the combination with Tamoxifen or chemotherapy as a first line palliative treatment is superi