It is known that amenorrhea, oligomenorrhea, irregular menstrual cycles, luteal insufficiency and infertility are frequently associated with hyperandrogenism. It has been reported in previous studies that the traditional herbal medicine, Shakuyaku-Kanzo-To(SKT) can lower high serum testosterone levels in oligomenorrheic or amenorrheic women, and that some of these sterile women conceive. SKT contains Shakuyaku (S) and Kanzo (K) in equal amounts. The main component of S and K is paeoniflorin and glycyrrhizin, respectively. This study was designed to investigate the mechanism in lowering serum testosterone (T) levels by SKT. Experiment I: Female Wistar rats were injected subcutaneously with 500 micrograms of testosterone propionate at the age of 2 days, becoming androgen-sterilized rats (ASR). Fifty-six-day-old ASR were given orally SKT (22.5, 45, 90 and 180 mg/kg body weight), S or K (11.25, 22.5, 45 and 90 mg/kg b.w.) in water through a tube every day for 2 weeks. Control ASR were given only water. Each group consisted of 10 rats. Serum total and free T levels in SKT and S groups were significantly lower than those in the controls, and these decreases were dose-dependent. Experiment II: Female Wistar rats were oophorectomized at the age of 60 days. From one week later they were given orally SKT (90 and 180 mg/kg b.w.), S or K (45 and 90 mg/kg b.w.) every day for 2 weeks. Control rats were given only water. Each group consisted of 11 rats. There was no change in serum T, LH and FSH levels in either groups. The results from experiment I and II suggest that SKT influences the T production by ovaries but not by adrenal glands. Experiment III: The minced tissues of one ovary obtained from proestrous Wistar rats were incubated with media containing Paeoniflorin(P), Glycyrrhetinic acid(GA) or Glycyrrhizin(GL) (GL) (1, 50 and 100 micrograms/ml, respectively, n = 5 in each group) for 270 minutes at 37 degrees C under an atmosphere of 95%O2 and 5%CO2. The T production by ovaries was significantly decreased in each treated group in comparison with the control, and this decrease was dose-dependent. However, the delta 4-androstenedione (delta 4-A) production by ovaries was increased in each treated group. The ratio of T to delta 4-A was significantly lower in each treated group than in the control. The estradiol(E2) production by ovaries in each treated group was not changed in comparison with the control.(ABSTRACT TRUNCATED AT 400 WORDS)