The identification of biological "markers" indicating distinctively different functions between preneoplastic and neoplastic as compared with normal cells has been a subject of intensive investigation, especially as additional technology becomes available. Although no distinct single biochemical marker is ubiquitous to the process of neoplasia or even to a single histogenetic type of neoplasm, a variety of histogenetic types of neoplasms in the human and experimental animals exhibit an extreme degree of marker or phenotypic heterogeneity. Particularly well studied are markers which occurred during the process of hepatocarcinogenesis in the rodent as well as in its final product, the hepatoma. Although phenotypic heterogeneity is characteristic of hepatocellular carcinomas in both the rat and mouse, some degree of predominant marker pattern(s) has become apparent. In multistage hepatocarcinogenesis in the rat a frequent but not completely ubiquitous marker is the enzyme gamma-glutamyltranspeptidase. In the mouse, although such markers have not been as extensively studied as in the rat, glucose 6-phosphatase is a predominant but not exclusive histochemical marker. Many preneoplastic lesions occurring during the stage of promotion exhibit reduced levels of enzymes of oxidative xenobiotic metabolism, but this pattern is not ubiquitous. Studies on the transcription of specific genes in mouse liver as well as preneoplastic and neoplastic lesions in this tissue further demonstrate the phenotypic heterogeneity characteristic of differentiated hepatocellular carcinomas. In general, current evidence supports the two theses that no single biologic marker or set of markers is uniquely characteristic of the preneoplastical and/or neoplastic phenotype and marker or phenotypic heterogeneity is by far the rule rather than the exception in hepatocarcinogenesis in the rodent and quite possibly in all histogenetic types of neoplasms in mammals.