Cytochemical analysis at the ultrastructural level was performed to characterize expression of catalase and glucose-6-phosphatase (G6Pase) activity as possible differentiation markers in oval cells proliferating during hepatocarcinogenesis induced in woodchucks by chronic infection with the woodchuck hepatitis virus (WHV) and additional treatment with aflatoxin B1 (AFB1). Oval cells from WHV-carriers treated with AFB1 showed two types of catalase-positive organelles: 1) microperoxisomes appearing as small strongly osmiophilic bodies corresponding to those present in biliary cells from control woodchucks, 2) peroxisomes with a hepatic staining pattern resembling those of mature hepatocytes but lacking a nucleoid. While in oval cells penetrating into the parenchyma a catalase-positive reaction product was restricted to rare microperoxisomes, in close vicinity to the portal tract about 30% of the oval cells produced peroxisomes with a hepatic staining pattern, indicating the existence of two different populations within the oval cell compartment. Peroxisomes with a hepatic staining pattern formed clusters and exhibited pleomorphism with marked variation in shape and size, the size sometimes coming up to that of normal hepatocellular peroxisomes. Serial sections revealed the complex organization of these peroxisomes. They consisted of several interconnected segments forming a peroxisomal reticulum. These findings are consistent with the hypothesis that a subpopulation of oval cells represents committed precursor cells capable of differentiating into hepatocytes. Activity of G6Pase was not demonstrable in this subpopulation of oval cells and became positive only in transitional cells. Differential expression of catalase and G6Pase activity in a stepwise fashion within the oval cell compartment appear to mark differentiation of oval cells into hepatocytes. Thus, elevated expression of catalase may be a useful early marker for the distinction of different subpopulations of oval cells committed to hepatic cell lineages before definitely changing their phenotype, whereas expression of G6Pase activity seems to begin later, accompanying morphological changes towards the phenotype of mature hepatocytes.