Central nervous system hormones have been linked to premenstrual syndrome (PMS) and beta-endorphin (beta-EP) is thought to be involved in the pathophysiology. We have tested the efficacy of the synthetic steroid Org OD 14 (tibolone) in the treatment of PMS. This prospective, randomized, placebo-controlled, double-blind cross-over study included 18 ovulatory women with PMS as ascertained by a visual linear analogue scale (VLAS). The women in each group received either 2.5 mg per day Org OD 14 (n = 9) or a multi-vitamin pill as placebo (n = 9) for 3 months. Treatments were then crossed over to a placebo for a further 3 months. VLAS ratings were evaluated at the end of each menstrual cycle throughout the study. Peripheral beta-EP concentrations were determined by radioimmunoassay on days 7 and 25 of each menstrual cycle. Changes in VLAS score and beta-EP concentrations from baseline were calculated and analysed by Student's paired t-test. Improvements in VLAS scores and beta-EP concentrations were evident during the second and third months of tibolone treatment. At the end of the third month, there was a significant improvement in VLAS scores of all symptom categories compared with pretreatment and placebo during treatment with tibolone (P < 0.05). Similar results were obtained in the first placebo group when switched to tibolone. Beta-EP concentrations were not significantly different between the study groups at the initial cycle (15.9 +/- 3.6 versus 17.2 +/- 2.3 pg/ml). The increase in beta-EP concentration was significantly greater on day 25 of the menstrual cycle in women treated with tibolone compared with baseline and placebo group (22.5 +/- 4.4 versus 15.9 +/- 3.6 and 17.2 +/- 2.3 pg/ml respectively, P < 0.05). Our data confirm the clinical efficacy of tibolone in PMS-related symptoms, as well as its effects on serum beta-EP concentrations in patients with PMS.