The purpose of the study was to determine the correlation between prognosis and chemotherapy induced amenorrhoea or elevated gonadotropin levels in node-positive breast cancer patients. Since we have previously found a better prognosis in patients with more profound leucopenia induced by adjuvant chemotherapy, we examined whether this effect was mediated through more efficient induction of amenorrhoea. The study population consisted of 126 premenopausal, primarily operable, node-positive breast cancer patients treated with cyclophosphamide, methotrexate and 5-fluorouracil (CMF) adjuvant chemotherapy at the Department of Oncology, Helsinki University Central Hospital between 1990 and 1993. 12 months after the beginning of adjuvant chemotherapy, the patients were divided into groups with respect to their menstrual function (regular menstruation, irregular menstruation or amenorrhoea). Information about menstruation status and serum concentration of follicle stimulating hormone (FSH) and oestradiol were recorded at 12 and 24 months from the beginning of adjuvant chemotherapy. Median follow-up time was 72 months. Women who experienced amenorrhoea or had irregular menstruation after chemotherapy had a significantly better 5-year disease-free survival (DFS) in univariate analysis than women who continued to menstruate (P = 0.02). Amenorrhoea and irregular menstruation were associated with a better DFS among patients with oestrogen receptor (ER) positive primary tumours (P = 0.007), whereas no such association was found in ER negative cases (P = 0.86). 5-year overall survival (OS) in univariate analysis was also better in patients who experienced amenorrhoea (81%) or who had irregular menstruation (90%) after chemotherapy as compared with patients with regular menstruation (68%; 81 versus 68%, P = 0.05). The serum FSH level did not correlate significantly with outcome irrespective of the cut-off point chosen. Nodal status, tumour size and menstruation status after chemotherapy were also significantly associated with DFS in a multivariate analysis. The menstruation status after chemotherapy lost its significance for OS in a multivariate analysis whilst the number of affected lymph nodes, tumour size and oestrogen/progesterone receptor status retained their impact. There was no association between the degree of leucopenia and induction of amenorrhoea by CMF. Chemotherapy-induced ovarian function suppression (amenorrhoea/irregular menstruation) after chemotherapy had a favourable effect on DFS in premenopausal breast cancer patients. The post-chemotherapy menstruation status is a clinically usable marker for sufficient endocrine effect of chemotherapy in ER/PR-positive patients in all premenopausal age groups. FSH level seemed to be a less reliable indicator of the castration effect of adjuvant chemotherapy in this study.