To investigate the effect of medrogestone on bone mineral density (BMD) and bone turnover under conditions of estrogen withdrawal, premenopausal women with endometriosis were treated with goserelin (Zoladex), combined with either placebo (group A, n = 12) or 10 mg medrogestone (Prothil, group B, n = 11) for six months, and followed for an additional six months. Lumbar spine BMD was measured at 0 and 6 month. Markers of bone turnover were serum bone alkaline phosphatase (sBAP) and osteocalcin (sOC) by ELISA, and urinary total pyridinoline (uPYD) and deoxypyridinoline crosslinks (uDPD) by HPLC. Patients in both groups had a similar and significant decrease in BMD after 6 months (4%, p < 0.01). The time course of changes in bone turnover, in contrast, was different in both groups. In group A, crosslink excretion increased from one month onwards, while no changes were seen in group B. In group A, sBAP levels rose during treatment, while in group B, this rise was delayed until treatment was terminated. Additionally, group B showed an initial suppression of sBAP and sOC. In both groups, sOC increased after treatment was discontinued. Medrogestone at 10 mg/d does not prevent lumbar bone loss in premenopausal women under estrogen deprivation. In the medrogestone add back group, the changes in bone turnover are compatible with low turnover bone loss,as ooposed to a state of high turnover seen in the unopposed goserelin group. This effect may be due to glucocorticoid receptor mediated actions of medrogestone on bone.