The effect of the clinically important anticonvulsant phenytoin (DPH) on hepatocarcinogenesis of male F344/NCr rats initiated with a single i.p. dose of N-nitrosodiethylamine (75 mg/kg b.w.) was studied. Beginning 2 weeks post-initiation, the rats received control diet or diet containing 500 or 1,500 ppm DPH or 500 ppm phenobarbital. At 52 weeks age, the incidences (and multiplicities, in units of tumors per tumor-bearing rat) of hepatocellular adenomas were 0%, 17% (1 +/- 0), 42% (1.8 +/- 0.8), or 67% (2.5 +/- 1.9) in rats exposed to N-nitrosodiethylamine alone, or the carcinogen followed by 500 ppm DPH, 1,500 ppm DPH, or 500 ppm phenobarbital, respectively. Between 53 and 79 weeks of age, 39% of rats receiving N-nitrosodiethylamine alone developed multiple (1.5 +/- 0.8) hepatocellular adenomas. A similar incidence (41%) occurred in the rats administered the carcinogen followed by 500 ppm DPH. The incidence of hepatocellular adenomas (88% and 89%) was significantly greater in rats exposed to N-nitrosodiethylamine followed by 1,500 ppm DPH or 500 ppm phenobarbital, respectively. Multiplicities of hepatocellular adenomas were significantly greater than the control value in rats fed 1,500 ppm DPH or 500 ppm phenobarbital (5.9 +/- 4.8 and 10.1 +/- 6.7, respectively), but not in the rats receiving 500 ppm DPH (2.3 +/- 1.6). No rats exposed to N-nitrosodiethylamine alone or the carcinogen followed by 500 ppm DPH developed hepatocellular carcinomas, while hepatocellular carcinomas occurred in 29% or 67% of the rats given 1,500 ppm DPH or 500 ppm phenobarbital, respectively, following initiation. Increases in hepatic CYP2B-mediated benzyloxyresorufin O-dealkylation activity in rats exposed to 500 and 1,500 ppm DPH for 2 or 23 weeks were approximately 50% and approximately 100%, respectively, of the maximal induction caused by 500 ppm phenobarbital. Thus, in the rat model, DPH enhanced N-nitrosodiethyl-amine-initiated hepatocarcinogenesis when administered at a dose causing maximal CYP2B induction.