Clinical pathology data can significantly contribute to the characterization of a disease process if suitable time points for sample collection are chosen and combined with the measurement of biochemical analytes that are sensitive and specific for damage to a potential target organ. Using a well-defined model for hepatotoxicity, we correlated histopathological lesions in the liver with changes in selected serum analytes. Groups of Fischer-344 rats were treated with carbon tetrachloride (280 mg/kg in corn oil) for 1, 2, 4, 6, 8 or 10 days. Subgroups were allowed to recover for 1, 5 or 8 days, at which time blood and liver specimens were collected. Histologically, necrosis was detected in livers from rats treated for 1 and 2 days and allowed to recover for 1 day. This was followed by generalized fatty change in animals treated for longer periods. The maximum severity of fatty change occurred 7-12 days (total experimental time). A sharp rise and fall (48 h) in cytosolic enzyme activities were seen in serum. This preceded gradual increases in all analytes measured which eventually peaked at 9-11 days (total experimental time). The pattern seen in biochemical analytes paralleled the development of marked fatty change. We discuss relationships between the histologic and biochemical findings and conclude that appropriate clinical biochemistry measurements in a toxicology experiment can provide valuable mechanistic information.