In a large group of unselected postmenopausal women, we investigated the predictive value of climacteric complaints for future bone loss. We reviewed data on 143 healthy, early postmenopausal women, who had participated in two placebo-controlled trials of the effects of different hormonal replacement regimens on bone mass, receiving either continuous estradiol valerate (E2V) and cyproterone acetate or sequential E2V and levonorgestrel, medroxyprogesterone acetate, or desogestrel. Follow-up examinations were done every 3 months, and 118 women (85%) completed the 2-year study. Bone mass was measured in the distal forearm (BMCarm) and the lumbar spine by photon and x-ray absorptiometry; bone turnover was estimated by measurements of plasma bone Gla protein, serum alkaline phosphatase, and fasting urinary hydroxyproline and calcium corrected for creatinine (FuHPr/Cr). Menopausal complaints were scored according to the Kupperman index and flush scores separately. At baseline, levels of serum estradiol and bone turnover parameters were independent of the Kupperman index and flush score. A slight negative relationship was seen between menopausal complaints and HPr/ Cr only (r = - 0.18; p < 0.05). In the placebo group there was no relationship between the spontaneous rates of bone loss during the 2 years and either the initial Kupperman index or flush score. The responses in both bone compartments to hormone replacement therapy (HRT) were also unrelated to the initial level of menopausal complaints. After 2 years the responses to HRT in bone turnover parameters and BMCarm were independent of menopausal complaints. We conclude that there is no evidence of a relationship between the severity of menopausal complaints and the rate of bone loss or the response to HRT. Women with severe menopausal complaints are a target group for HRT because of the symptom relief, but those requiring HRT for prevention of osteoporosis must be identified by other means.