Female F344/N rats were initiated with DEN (10 mg/kg) 24 h after a 70% partial hepatectomy. Groups of 10 rats were fed (a) AIN, group-1; (b) PD, group-2; or (c) NIH, group-3, for 1 week after initiation and were then fed NIH plus the promoting agent PB at a level of 0.05% in the diets for 6 months. Other groups were fed NIH for 1 week after initiation and then NIH without PB (group-4), AIN + PB (group-5), AIN without PB (group-6), PD + PB (group-7), or PD without PB (group-8) for 6 months. The numbers and volume percentages of AHF were quantified by stereologic methods from frozen serial sections, stained consecutively for GGT, ATPase, and G6Pase. For the groups fed different diets during the 1st week after initiation, the numbers and volume of AHF were significantly greater in group-2 than in groups 1 or 3. The numbers of AHF were significantly less in group-3 than in group-1. The numbers and volume of AHF were significantly greater in groups fed PB during the promotion phase, except in the case of group-7, whose focal volume did not differ from groups 6 or 8. Group-3 had significantly greater numbers of AHF than groups 5 and 7. These findings can be explained by the hypothesis that the NIH diet contained factors that acted synergistically with PB to enhance tumor promotion. The mean focal volume of both GGT positive and ATPase negative foci was significantly greater in group-5 than in all other groups; this indicates that the AIN + PB regimen selectively promoted the growth of a subpopulation of AHF. These findings show that alterations in the composition of diets fed during hepatocarcinogenesis significantly alter the effects of specific chemical agents acting during the stages of initiation and promotion in hepatocarcinogenesis.