Recently, we used human hepatocytes in primary culture to study the effects of inducers of glutathione-S-transferases (GSTs) in the expectation that information obtained can be used to predict the value of particular inducers for use in the chemoprevention of cancer and other toxicities. However, in vitro human studies cannot readily be confirmed by studies in vivo. This problem does not arise in experimental animals. In the current studies, the response of male rat hepatocytes in primary culture to the following inducers of GST isoenzymes has been determined: 3-methylcholanthrene (MC); phenobarbital (PB); 1,2-dithiole-3-thione and its 5-(2-pyrazinyl)-4-methyl derivative, oltipraz (OPZ), and the results have been compared with induction obtained in livers of MC- and OPZ-treated rats. Each type of inducer was found to elicit a different response. In vitro, phenobarbital increased messenger RNA (mRNA) levels of subunits 1b and 3 after 12 and 72 hours, respectively; MC had a rapid effect on GST alpha class mRNAs (bringing about increase after only 2 hours of treatment), increased subunit 7 mRNA slightly, and had no effect on mu class mRNAs; dithiolethiones induced both subunit lb and 7 mRNAs after 4 hours and, to a much lower extent, subunit 3 mRNA after 72 hours. In vivo, MC induced significantly both subunit lb and 7 mRNAs whereas OPZ increased significantly subunits lb, 3 and 7 mRNA levels, and to a lower extent those of subunit 2, after 3 days and beyond to at least 5 days of treatment. Results obtained in mRNA studies were confirmed by high-pressure liquid chromatography (HPLC) analysis of GST subunits. HPLC also showed an induction of subunit 10 at the protein level of which the mRNA was not analyzed. Our results show that rat hepatocytes in primary culture prove to be a good model for the effect of inducers on both the expression of GST mRNA and protein levels in the rat liver in vivo. The demonstration of this good correlation in the rat with respect to increases gives support for the use of human hepatocytes for predictive studies of chemoprotection in human pharmacology.