The Peto test is the standard method of analysis used in carcinogenicity st
udies to compare tumor incidence in groups of animals. It assumes that tumo
rs are either instantly fatal or have no effect on mortality and requires a
judgement of the lethality of each tumor. To avoid this requirement, param
etric multi-state models have been proposed. In addition these allow estima
tion of tumor onset and mortality rates. This paper considers two such mode
ls and presents a modification. It is shown that the modified models provid
e a better fit to carcinogenicity data and simulated data are used to show
that the modified models provide a modest increase in test power relative t
o the Peto test.