Understanding the mechanism of carcinogens present in the workplace and the environment has become increasingly important because of the use of mechanism-based approaches to extrapolate results from laboratory animals to humans. There is great uncertainty in interspecies extrapolation because the dose levels used in animal studies are usually much greater than doses resulting from human exposure. High dose levels are required in animal carcinogenesis studies because of the low sensitivity of the bioassay. The uncertainty in the extrapolation is especially great for non-genotoxic carcinogens. This is because non-genotoxic carcinogens are likely to have dose-response curve that are not linear and that include a threshold. Furthermore, they could induce cancer in animals by a mechanism that is not applicable to humans. Knowledge of the mechanism of these carcinogens allows the development of molecular and biological markers for their activity. One of these biomarkers is the hypomethylation of DNA and protooncogenes. Biomarkers could be used to determine whether the mechanism applies to humans and to better define the dose-response relationship extending it to exposures/dose levels not applicable to carcinogenesis bioassay. Lung and mouse liver non-genotoxic carcinogens will be investigated, since the lung is a major site of occupational-related cancer and mouse liver is a major target organ in carcinogenesis bioassays.