In vitro systems of high biological organization, e.g. containing intact hepatocytes, have been considered as more reliable for metabolic studies and in vivo predictions of toxicokinetics than subcellular systems. For this reason, the kinetics and metabolism of low-molecular-weight volatile chemicals, i.e. head space elimination of substrate and formation of metabolites, were compared in liver S9 and in liver slices. Two substrates, toluene and n-hexane, were used as they represent differences in metabolic pathways and physical-chemical properties. The two systems responded similarly to diethyldithiocarbamate inhibition and acetone inhibition/induction of cytochrome P-450 mediated metabolism. In contrast to liver S9, liver slices did not respond adequately to phenobarbital induction raising the question of substrate availability for P-450 enzymes in liver slices. Liver slices appeared to be superior to liver S9 when specific metabolic pathways involving phase II enzymes were investigated. However, liver S9 seemed to be at least as good as liver slices for estimation of total metabolic rate constants (Km, Vmax) as a basis for in vivo predictions. As the head space liver S9 system is faster and easier to operate than liver slices, it is a promising screening tool for the metabolism of volatile compounds and metabolic interactions.