The effects of phenobarbital (PB) on hepatocellular carcinogenesis in three strains of nitrosamine-initiated infant male mice were evaluated. Fifteen-day-old C57Bl/6NCrlBR (C57Bl), C3H/HeNCr1BR (C3H) and B6C3F1 mice were treated with a single i.p. injection of either diethylnitrosamine (DENA) (5 micrograms/body wt), dimethylnitrosamine (DMNA) (5 micrograms/body wt) or saline. One-half of the treated mice received PB via the drinking water (500 mg/l) for 24 weeks. The remaining treated mice were given deionized drinking water. Mice were killed at 28 weeks of age and hepatic lesions were evaluated. Only animals that received DENA or DMNA exhibited tumors. C3H mice treated with DENA + PB demonstrated a significant increase in hepatic adenoma number compared to C3H mice exposed to DENA only. Conversely, B6C3F1 males treated with DENA + PB exhibited a significant decrease in the number of hepatic adenomas compared to B6C3F1 males treated with DENA alone. No change was noted in adenoma size in B6C3F1 mice treated with DENA + PB from those receiving DENA only. Chronic PB exposure of C57Bl males previously treated with DENA had no effect on hepatic adenoma number or size. C3H mice treated with DMNA + PB displayed an increase in both adenoma size and adenoma number compared to C3H mice receiving DMNA only. Similarly, in B6C3F1 mice, PB treatment increased both the adenoma incidence and adenoma number in DMNA initiated mice. PB had no effect on hepatic adenoma incidence or number in DMNA-treated C57Bl mice. These data suggest that the ability of PB to promote hepatic tumorigenesis in the 15-day-old initiated mouse is dependent on both the strain of the mouse and the initiating chemical carcinogen.