Peroxisome proliferators (PPs), such as diethylhexylphthalate (DEHP), constitute a diverse class of chemicals with many therapeutic, industrial and environmental applications. In rodents, PPs are nongenotoxic hepatocarcinogens, raising concerns regarding the potential of PPs to harm human health. However, humans differ from rodents in their response to PPs and the weight of evidence supports the supposition that PPs do not pose a carcinogenic risk to humans. The effects of PPs in the rodent are mediated by peroxisome proliferator activated receptor alpha (PPARalpha). PPARalpha predominates in the liver whereas another isoform PPARgamma predominates in adipose tissue and in the immune system. This tissue-specific pattern of PPARalpha expression is consistent with a role for PPARalpha but not PPARgamma or PPARbeta in PP-induced rodent hepatocarcinogenesis. Humans, marmosets and guinea-pigs appear refractory or less responsive to the adverse liver effects of PPs. However, humans give a therapeutic response to the fibrate PPs via an alteration in lipid metabolism mediated by PPARalpha. Such marked species differences may be explained by quantity of PPARalpha and/or the quality of the PPARalpha-mediated response. The lower expression of full-length functional PPARalpha in humans could be attributed to the presence of a truncated, inactive form of PPARalpha, which appears to be present in most individuals examined to date. In addition, there are species differences in sequence and responsiveness of the acyl CoA oxidase (ACO) gene promoter, suggesting that even in the presence of sufficient PPARalpha, the human equivalent of rodent genes associated with peroxisome proliferation may remain inactive.