Liver parenchymal cells (hepatocytes) of human organ donors were isolated using a two-step collagenase perfusion technique. The average viability of the freshly isolated liver parenchymal cells, as judged by trypan blue exclusion, was 82% (SD = 7%; n = 6). The inter-individual differences in the determined enzyme activities were less than a factor of 7.5, despite the different sexes and ages of the donors. Freshly isolated parenchymal cells (PC) were cryopreserved using a computer-controlled freezing protocol. After thawing, cryopreserved cells had a mean viability of 57% (SD = 18%; n = 6). The activities of xenobiotic metabolizing enzymes in freshly isolated and cryopreserved cells were compared using PC from two donors. The enzyme activities of phenol sulfotransferase, 1-naphthol UDP-glucuronosyltransferase and microsomal epoxide hydrolase were well maintained after thawing (87-117% of activities in freshly isolated cells), whereas the activities of glutathione S-transferase, monitored with the broad spectrum substrate 1-chloro-2,4-dinitrobenzene, and the major broad spectrum cytosolic epoxide hydrolase were moderately but markedly reduced after cryopreservation (34-64% and 45-89% of levels in fresh cells, respectively). The decrease of both activities was dependent on the viability after thawing. When cryopreserved cells were purified by a Percoll centrifugation after thawing, the viability was increased from 62 to 92% for cells from one of the donors and from 88 to 98% for PC for the other donor. Subsequently the activity of glutathione S-transferase in Percoll-purified PC from the two donors was increased to 71 and 96% of levels in freshly isolated cells. It is concluded that the use of cryopreserved liver parenchymal cells of humans and other species represents a valuable tool in predicting which animal species best represents humans in hepatic metabolism and therefore should be the preferred species for investigations of metabolism and metabolism-dependent toxicities.