The ability of methyl-deficient, amino-acid-defined diets to produce enzyme-altered foci was quantitatively determined in the livers of rats treated both with and without an initiating dose of diethylnitrosamine (DEN). Male weanling F-344 rats were fed a complete, amino-acid-defined diet for 1 week. They were then injected i.p. with a single dose of DEN (20 mg/kg body weight) and fed the complete diet for an additional week. Forty animals in each dose group were then maintained for 5-38 weeks on the complete diet (diet 1) or one of the three methyl-deficient diets customarily used in this laboratory: diet 2, devoid of methionine and choline; diet 3, devoid of methionine only; and diet 4, devoid of choline only. In diets 2 and 3, methionine was replaced by equimolar amounts of its metabolic precursor, DL-homocystine. Ten animals per group were killed 8, 12, 17, 24 and 41 weeks after DEN initiation. For 2 weeks prior to being killed, each group was maintained on the complete diet to minimize the histological abnormalities due to acute toxicity of the diets. Serial sections of the livers were obtained, stained sequentially for gamma-glutamyltranspeptidase, ATPase and glucose-6-phosphatase, and the quantitation of the focal lesions scored by these markers was carried out by quantitative stereology. The results indicated that, regardless of the enzyme marker(s) examined, there was a general correspondence between the volume and number of altered hepatic foci (AHF) formed and the previously described tumor-promoting activities of each diet. Thus, while all DEN-treated groups contained significant numbers of AHF 24 weeks after initiation, only the diet-2-fed animals displayed such foci at 8 weeks. Similarly, among the uninitiated rats, only those fed diet 2 exhibited the presence of AHF throughout the experimental period. Interestingly, the livers of uninitiated, choline-deficient rats showed a small number of AHF at 24 and 42 weeks; these foci were not observed at all in the corresponding DEN-untreated animals fed diet 3, deficient in methionine only. The results provide evidence that the carcinogenic effects of the methionine- and choline-deficient diet result more from its strongly promoting effect than from any initiating activity by the diet.