The frequency and severity of ankylosing spondylitis (AS) show a male preponderance, and androgenic steroids have been implicated in its etiology. Some reports have indicated that serum androgen levels are slightly elevated relative to estrogen levels in patients with AS as compared to controls. In more recent studies, however, serum testosterone, 17 beta-estradiol, and androstenedione levels did not significantly differ between AS patients and controls. Moreover, testosterone levels measured directly in serum can be spuriously elevated, especially in patients using phenylbutazone. Elevated serum levels of the adrenal steroids 17 alpha-hydroxyprogesterone and dehydroepiandrosterone (DHEA) sulfate have been found in patients with AS. These elevations might be explained by partial 11 beta- or 21-hydroxylase deficiencies, but may also be secondary to an enhanced stress response. In vitro studies as well as studies in animals and humans indicate that DHEA enhanced, and 17 beta-estradiol and progesterone inhibit, the cell-mediated immune response, which may play a role in the pathogenesis of AS. Oral estrogen therapy in female patients and human chorionic gonadotrophin injections in male patients with AS, increased the 17 beta-estradiol/testosterone ratio and resulted in a moderate clinical improvement. In conclusion, serum testosterone levels are not elevated in patients with AS. Therefore testosterone probably has no role in the perpetuation of long-standing AS and provides no basis for antiandrogenic treatment. Cross-sectional case-control studies, however, cannot clearly distinguish etiological factors from secondary disease effects, especially when blood sampling occurs many years after the onset of AS. Consequently, the role of sex steroids in the pathogenesis is still insufficiently elucidated.