Several studies have demonstrated that bisphenol A (BPA) exhibited weak estrogenic activity in the 3-day uterotrophic assay using ovariectomized (OVX) and immature rats (Toxicol. Lett. 115 (2000) 231; Regul. Toxicol. Pharmacol. 32 (2000) 118; J. Toxicol. Sci. 26 (2001) 111) and BPA also possessed anti-androgenic activity in in vitro yeast based assays (J. Endocrinol. 158 (1998) 327). To investigate anti-androgenic effects of BPA. a rodent Hershberger assay was carried out using immature Sprague-Dawley male rats. An androgen agonist, testosterone (0.4 mg/kg per day), was administered for 7 consecutive days by subcutaneous (s.c.) injection as a positive control. Additionally, a pure androgen antagonist, flutamide (1, 5. 10 mg/kg per day. oral) was co-administered with testosterone (0.4 mg/kg per day s.c.). BPA was also administered orally with or without testosterone (0.4 mg/kg per day, s.c.) for 7 consecutive days. In the testosterone treated groups, glans penis, seminal vesicles, ventral prostate, and levator ani plus bulbocavernosus muscles (LABC) weights were significantly increased compared with control. However. flulamide dose-dependently inhibited the testosterone-induced re-growth of seminal vesicles, ventral prostate, and LABC, with a significant decrease at flutamide 1.0 mg/kg and above (P<0.05). Serum LH levels were also significantly increased (5 mg/kg and above, P<0.05), but no changes in serum testosterone levels. In contrast, BPA had no effects on the re-growth of seminal vesicles, ventral prostate and LABC induced by testosterone, and no significant differences were observed in serum LH and testosterone levels. In summary, the Hershberger assay could be a sensitive method for detecting androgenic or anti-androgenic chemicals, but BPA did not exhibit any androgenic or anti-androgenic activities in Hershberger assay.