Phenobarbital (PB) is an effective growth stimulator of hepatic hyperplastic nodules developed with diethylnitrosamine and 2-acetylaminofluorene plus partial hepatectomy (the Solt-Farber model), but it does not apparently stimulate the growth of preneoplastic lesions produced with aflatoxin B1 (AFB). Some studies have suggested a correlation between the induction of specific cytochrome P450 enzymes and the tumor promoting effects produced by repeated treatment with PB. To examine this hypothesis further, hepatic hyperplastic nodules were produced with AFB (10 ip doses of AFB, 150 micrograms/kg/day, followed by partial hepatectomy) or by a modified Solt-Farber protocol (DEN/AAF), and the effects of PB on nodule growth and expression of cytochrome(s) P450 2B1 and/or P450 2B2 (P450 2B1/2) were determined. Both treatment protocols (without PB) produced multiple, large nodules within 10-17 weeks of carcinogen administration. These nodules stained intensely for glutathione S-transferase p (GST-p; GST7-7) and gamma-glutamyl transpeptidase (GGT) and weakly for P450 2B1/2. Pentoxyphenoxazone dealkylation activity was decreased to less than 50% of the surrounding tissue levels in both types of nodules. PB treatment of animals with DEN/AAF-induced nodules greatly increased P450 2B1/2 expression in surrounding tissues, whereas most, but not all, nodules were not inducible. Pentoxyphenoxazone dealkylation was increased 31- to 35-fold in surrounding tissue, but it was increased only 2-fold in pooled nodular tissue, relative to untreated control liver. In contrast to the DEN/AAF group, immunohistochemical staining and pentoxyphenoxazone dealkylation in the AFB group demonstrated that P450 2B1/2 was equally inducible in nodular and surrounding tissues. Short-term treatment (5 days) with PB produced a 2-fold increase in the number and total area of GGT-positive nodules in the DEN/AAF group, but it had no significant effect on the number, size distribution, or total area of GGT-positive nodules in the AFB group. All large GGT-positive nodules in the DEN/AAF group were nonresponsive to induction of P450 2B1/2, whereas all of the GGT-positive nodules which were responsive to P450 2B1/2 induction by PB in this group were relatively small. The size and area of AFB-induced GGT-positive nodules was not affected by PB treatment, and P450 2B1/2 in all of these nodules was inducible by PB. Although a causal, inverse relationship between the responsiveness of nodules to PB induction of P450 2B1/2 and their reaction to PB growth stimulation cannot be firmly established, these data are consistent with such a hypothesis.