Three protocols were used to determine the effects of dietary selenium concentration on the development of gamma-glutamyl-transpeptidase (GGT)-positive foci and hepatocellular carcinoma induced by either diethylnitrosamine (DEN) or N-acetylaminofluorene in rats. In the first experiment, foci were induced by a carcinogenic dose of DEN (100 mg/kg body weight, p.o.) at 20-22 h after two-thirds partial hepatectomy. One wk after DEN administration, during which time 0.1 ppm (representing a control level), 3.0, or 6.0 ppm selenium as Na2SeO3 was fed for 8 or 16 wk, at which time focal analysis was conducted using quantitative stereology. The results demonstrated that 3.0 and 6.0 ppm dietary selenium, initiated 1 wk following carcinogen administration, decreased focal growth rate without affecting the number of GGT foci compared to 0.1 ppm selenium. Decreased focal growth was temporary and reversible with 6.0 ppm selenium which may be related to chronic selenosis observed after 16 wk of 6.0 ppm selenium feeding. A second experiment involved a noncarcinogenic dose of DEN (25 mg/kg body weight, p.o.), then 0.1 or 6.0 ppm selenium feeding for 8 wk, followed by 0.05% phenobarbital (PB), a liver tumor promoter in a diet containing 0.1 ppm selenium. Analysis of GGT foci at 5 or 8 wk of PB feeding indicated that 6.0 ppm selenium caused a trend towards an increase in the number of foci/cm3 of liver and mean focal volume and a significant increase in GGT focal volume as a percentage of liver volume by 8 wk of PB feeding. Thus, high dietary selenium concentrations prior to PB enhance the tumor-promoting ability of PB. In a third experiment, using male Fischer 344 rats (150 g), 0.1 or 6.0 ppm selenium was fed concurrently with 0.02% AAF which was fed in a cyclic regimen. After 4 cycles, where 1 cycle equalled 4 wk of AAF, followed by 1 wk of control diet (0.1 ppm selenium), 6.0 ppm selenium significantly decreased the mean focal volume and focal volume as a percentage of liver volume, while not affecting the number of foci/cm3 of liver, again indicating a selenium effect on focal growth while not affecting the number of "preneoplastic" lesions in the liver. Six ppm selenium feeding after AAF treatment had no effect on the percentage of incidence of hepatocellular carcinoma (100%) but did cause a significant decrease in the percentage of liver volume occupied by macroscopic subcapsular liver lesions compared to 0.1 ppm selenium.(ABSTRACT TRUNCATED AT 400 WORDS)