Background The polycystic ovary syndrome ( PCOS) affects approximately 6-10% of women of reproductive age and is characterized by chronic anovulation and hyperandrogenism. However, a comprehensive understanding of the mechanisms that dictate androgen overproduction is lacking, which may account for inconsistencies between measures of androgen excess and clinical presentation in individual cases. Methods A rat model of PCOS was established by injecting dehydroepiandrosterone sulfoconjugate (DHEAS) into pregnant females. Rats were administered with DHEAS (60 mg/kg/d) subcutaneously (s.c.) for all 20 days of pregnancy (Group A), or for the first 10 days (Group B), or from day 11 to day 20 (Group C). Controls were administered with injection oil (0.2 ml/day) s.c. throughout pregnancy (Group D). The litter rate, abortion rate, and offspring survival rate in each group were recorded. Serum androgen and estrogen were measured and the morphological features of the ovaries were examined by light and electron microscopy in the offspring of each group. Results We found that rats injected with DHEAS throughout pregnancy (group A) lost fertility. Rats injected with DHEAS during early pregnancy (group B) exhibited more serious aberrations in fertility than both Group C, in which rats were injected with DHEAS during late pregnancy (P