Sixteen postmenopausal women receiving conjugated equine oestrogens 1.25 mg/day, continuously, were randomly allocated to add dydrogesterone 20 mg/day for 12 days each calendar month for 3 months and then 10 mg/day in an identical fashion for a further 3 months, or to receive the dydrogesterone doses in reverse sequence. The effects of the two dydrogesterone doses on endometrial histology, vaginal bleeding, and the symptomatic and psychological status were compared. Endometrial samples were obtained around day 10 of progestogen addition. Dydrogesterone, 20 mg, induced uniform, late secretory transformation in all samples; with 10 mg one sample showed mixed early and late secretory features and another demonstrated late secretory changes associated with atypical hyperplasia. Both dydrogesterone doses induced an acceptable withdrawal bleed; most bleeding episodes were 'spotting' or normal in amount, and heavy bleeding was reported infrequently. There was one episode of breakthrough bleeding. There were no differences in bleeding patterns between the two dose regimens. Anxiety, and the physical and psychological status were significantly improved after 3 months of therapy. Significant benefits on depression were observed less clearly. There were no differences between the two dydrogesterone doses on anxiety, depression and the physical and psychological status, and, overall, the addition of the progestogen did not antagonize oestrogen benefits.