[en] Background: During the last two decades of the 20th century, hormone replacement therapy (HRT) has been considered as the sole pharmacological approach for counterbalancing or mitigating the effects of estrogens deprivation in post-menopausal women. Subsequently, HRT has been widely recommended for the management of chronic diseases occurring, in women during the second half of their life. The overall risk/benefit ratio of estrogens has been recently reassessed in the light of long-term prospective studies failing to demonstrate the expected benefit of HRT on cardiovascular diseases incidence. Osteoarthritis (OA) is one of the chronic conditions for which HRT has been suggested to provide beneficial outcomes. Results: The presence of estrogen receptors in human cartilage is no longer debated. However, cellular or animal models of OA do not provide an unequivocal pathway for the influence of gonadal steroids on cartilage. Similarly, studies attempting to correlate serum or urinary levels of sex steroids to the onset or progression of OA gave conflicting results. No randomized, prospective, controlled trial was designed to specifically assess the impact of hormone replacement therapy on symptomatic or structural progression of OA. Large-scale observational studies or trials designed to assess other potential benefits of estrogens suggest that HRT use does not provide symptomatic relief in OA but may interfere with its long-term structural progression, particularly in the lower limbs. Conclusion: Based on the recent results of the Women Health Initiative suggesting that HRT health risks may outweigh benefits, one can hardly recommend, with the current level of evidence, HRT as a first-line treatment against progression of OA. (C) 2003 OsteoArthritis Research Society International. Published by Elsevier Science Ltd. All rights reserved.