The modifying effects of concurrent treatment with high or low doses of butylated hydroxyanisole (BHA) on wide-spectrum carcinogen-induced carcinogenesis were studied in male F344 rats. Groups of 20 animals were treated with 2 or 0.04% BHA for 24 weeks. Starting 2 weeks after the commencement of BHA treatment, they were given s.c. injection of 50 mg/kg body weight 3,2'-dimethyl-4-aminobiphenyl (DMAB) once a week, i.g. administrations of 200 mg/kg body weight 2,2'-dihydroxy-di-n-propylnitrosamine (DHPN) once every 2 weeks, or i.p. injection of 15 mg/kg body weight N-methylnitrosourea (MNU) once every 2 weeks for 22 weeks. Further groups of rats were treated with DMAB, DHPN, MNU, or 2 or 0.04% BHA alone. All surviving animals were killed 24 weeks after the beginning of the experiment and the target organs examined histopathologically. The BHA treatment dose-dependently decreased the incidence of DMAB-induced liver preneoplastic lesions but was associated with significant tumor induction in the forestomach (papillomas, 40%, P less than 0.01) and urinary bladder (papillomas, 53%, P less than 0.001; carcinomas, 80%, P less than 0.001), where no lesions were observed in the group given only DMAB. Concurrent administration of 2% BHA also significantly inhibited the development of alveolar hyperplasia (P less than 0.001) of the lung in DHPN-treated animals, while enhancing induction of forestomach papillomas (P less than 0.05) and simple hyperplasia in the urinary bladder. Neither MNU nor 2% BHA alone induced forestomach carcinoma or papillary or nodular hyperplasia (PN hyperplasia) in the urinary bladder. However, these lesions were observed in 100% (P less than 0.001) and 55% (P less than 0.001) of animals respectively, receiving the two compounds in combination. These results demonstrated that concurrent treatment with BHA not only inhibits but can also strongly enhance carcinogenesis depending on the organ, irrespective of whether the carcinogens act directly or require metabolism. The finding that BHA potently modified carcinogenesis at 0.04% in diet, 1/50 of the carcinogenic dose, suggests that actual dietary levels close to the human situation might play a significant role in tumor development in man.