The potential of polycyclic aromatic hydrocarbons (PAHs) to modulate microsomal epoxide hydrolase activity, determined using benzo[a]pyrene 5-oxide as substrate, in human liver, was evaluated and compared to rat liver. Precision-cut liver slices prepared from fresh human liver were incubated with six structurally diverse PAHs, at a range of concentrations, for 24h. Of the six PAHs studied, benzo[a]pyrene, dibenzo[a,h]anthracene and fluoranthene gave rise to a statistically significant increase in epoxide hydrolase activity, which was accompanied by a concomitant increase in epoxide hydrolase protein levels determined by immunoblotting. The other PAHs studied, namely dibenzo[a,l]pyrene, benzo[b]fluoranthene and 1-methylphenanthrene, influenced neither activity nor enzyme protein levels. When rat slices were incubated under identical conditions, only benzo[a]pyrene and dibenzo[a,h]anthracene elevated epoxide hydrolase activity, which was, once again accompanied by a rise in protein levels. At the mRNA level, however, all six PAHs caused an increase, albeit to different extent. In rat, epoxide hydroxylase activity in lung slices was much lower than in liver slices. In lung slices, epoxide hydrolase activity was elevated following exposure to benzo[a]pyrene and dibenzo[a,l]pyrene and, to a lesser extent, 1-methylphenanthrene; similar observations were made at the protein level. At both activity and protein levels extent of induction was far more pronounced in the lung compared with the liver. It is concluded that epoxide hydrolase activity is an inducible enzyme by PAHs, in both human and rat liver, but induction potential by individual PAHs varies enormously, depending on the nature of the compound involved. Marked tissue differences in the nature of PAHs stimulating activity in rat lung and liver were noted. Although in the rat basal lung epoxide hydrolase activity is much lower than liver, it is more markedly inducible by PAHs.