Combustion sources emit soot particles containing carcinogenic polycyclic organic compounds which are mutagenic in short-term genetic bioassays in microbial and mammalian cells and are tumorigenic in animals. Although soot is considered to be a human carcinogen, soots from different combustion sources and fuels contain variable quantities of organic tars of somewhat different composition. This presents a serious problem when assessing risk since it is not possible to collect human cancer risk data for all the various possible exposures to combustion emissions. Furthermore it is impossible to collect human cancer data for exposures to emissions from new technologies which have not previously been in the environment. The use of chronic animal testing for each new combustion technology developed would be prohibitively expensive and time consuming. This problem has motivated us to develop and test a comparative potency risk assessment methodology which would utilize short-term mutagenesis and tumorigenesis data.