Because there is usually more extensive toxicity, metabolism, and phar
macokinetic information for pharmaceuticals as opposed to environmenta
l agents, including pesticides, the argument has been made that carcin
ogenicity testing in two rodent species may not have been necessary fo
r carcinogenicity testing of pharmaceuticals. On the basis of numerica
l data only, it may be argued that carcinogenicity testing of pharmace
uticals in one species, typically the rat, is sufficient to identify p
otential human carcinogens. The argument that testing in a second spec
ies, typically the mouse, is redundant overlooks the value added by th
e second species carcinogenicity study. Bioassay data from the second
species allows balance and perspective in evaluating the observed effe
cts, and this is especially critical when there is a marginal, questio
nable, or inconclusive response in one species. Utilization of two spe
cies for carcinogen identification is the principal means for identify
ing trans-species carcinogens-those mostly likely to be carcinogenic i
n humans. Given that neither rat nor mouse are ideal surrogates for hu
mans, concordant data from both species strengthens the ability to ext
rapolate findings to humans. We believe that testing in two species sh
ould continue to be the default approach used for carcinogen hazard id
entification whenever scientifically indicated until such time that ac
ceptable and suitable alternatives are available. To utilize only one
species for this important means of protecting human health is prematu
re at this time.