1. The aims were to refine experimental conditions (using 76 human hepatocyte preparations) in terms of the selection of enzyme inducers and their optimal concentration, the treatment duration with inducers and the choice of specific cytochrome P450 isoform(s) probes to optimize the use of primary hepatocytes for predicting the potential induction by new chemical entities of cytochrome P450 isoforms in vivo in man. 2. In the absence of any inducer, basal cytochrome P450 isoform(s)-mediated activities decreased to 20% of their initial activity (end of the seeding period) by 72-96 h. In contrast, UGT-dependent enzyme activities remained at a constant level (+/- 20%) up to the fifth day of culture. 3. Beta-naphthoflavone, at an optimal concentration of 50 microM and after a 3-day treatment, specifically and potently induced 7-ethoxyresorufin (10.4 +/- 10.4-fold, n = 74) and phenacetin (6.6 +/- 6.4-fold, n = 60) O-deethylation processes, markers for CYP1A1 and CYP1A2 isoforms respectively. Only a 2-fold increase was noted following treatment with 2 mM phenobarbitone, whereas dexamethasone and rifampicin had no effect at all. 4. A 3-day treatment of human hepatocytes with 50 microM dexamethasone was associated with a major induction of both coumarin 7-hydroxylation (9.4 +/- 11.4-fold, n = 49) and nifedipine dehydrogenation (4.7 +/- 3.8-fold, n = 61), markers for CYP2A6 and CYP3A4 respectively. Phenobarbitone, however, exhibited a broad but moderate inducing effect on 7-ethoxyresorufin (2.2 +/- 1.5-fold, n = 55) and phenacetin (1.7 +/- 0.9-fold, n = 54) O-deethylation, coumarin 7-hydroxylation (3.9 +/- 9.2-fold, n = 50) and nifedipine dehydrogenation (2.1 +/- 2.0-fold, n = 47). 5. Km obtained for the different cytochrome P450 isoform substrates in untreated hepatocytes were in the same range of magnitude that those determined on human hepatic microsomal fractions. Enzyme induction processes were characterized by a large increase in apparent Vmax whereas apparent Km were not affected. 6. These studies demonstrate that human hepatocytes in primary culture can respond specifically and quantitatively to model inducers. This in vitro system offers a useful approach to study the regulation of human hepatic biotransformation activities and should facilitate the demand for a reproducible method for addressing cytochrome P450 induction.