Previous reports have shown that pancreatic cancer was induced preferentially in male versus female azaserine-treated rats. This study was designed to determine the importance of estrogen and testosterone in this phenomenon. Fischer (F344) rats received a single injection of azaserine (30 mg/kg) at 21 days of age. At 28 days of age, they were weaned and divided into 12 groups of 9–10 rats as shown below. Surgery (castration or sham operation) was performed at 4 weeks of age. All drugs (estradiol, the antiestrogen tamoxifen, testosterone propionate and/or the antiandrogen flutamide) were administered, starting at weaning, in 3-week timedrelease pellets until autopsy. Rats were killed 4 months after the administration of azaserine. The pancreas was weighed and prepared for quantitative histologic analysis of atypical acinar cell nodules (AACNs) which are putative preneoplastic lesions. Both number and size of AACNs were analyzed. In intact female rats, AACN burden was smaller than in intact males (P < 0.05). Ovariectomy increased the AACN burden (P < 0.05), while estradiol or tamoxifen treatments to ovariectomized females restored the burden to control levels (P < 0.05). Testosterone with tamoxifen treatment to ovariectomized females led to a significant increase in AACN burden over control values. In intact male rats, orchiectomy decreased the AACN burden (P < 0.05). In orchiectomized rats, testosterone treatment slightly increased the AACN burden, flutamide treatment alone increased this parameter (P < 0.05) but flutamide with estradiol decreased the AACN burden (P < 0.01). These data strongly support the hypothesis that sex steroids play a major role in the higher incidence of pancreatic cancer hi male versus female rats.