Molecular epidemiology has significant potential in preventing cancer and other diseases caused by environmental exposures (related to lifestyle, occupation, or ambient pollution). This approach attempts to prevent cancer by incorporating laboratory methods to document the molecular dose and preclinical effects of carcinogens, as well as factors that increases individual susceptibility to carcinogens. Recently we have carried out validation studies of biologic markers such as carcinogen--DNA and carcinogen--protein adducts, gene and chromosomal mutations, alterations in target oncogenes or tumor suppressor genes, polymorphisms in putative susceptibility genes (individual P450s, glutathione transferase M1), and serum levels of micronutrients. This research involves adults, infants, and children exposed to varying levels of carcinogens, as well as cancer cases and controls. On a group level, dose-response relationships have frequently been seen between various biomarkers and environmental exposures such as polycyclic aromatic hydrocarbons, cigarette smoke (active and passive), and ambient indoor and workplace air pollution. However, there is significant interindividual variation in biomarkers that appears to reflect a modulating effect on biomarkers (hence potential risk) by genetic and acquired susceptibility factors. Ongoing retrospective and nested case-control studies of lung and breast cancer are examining the association between biomarkers and cancer risk. Results of these studies are encouraging; they suggest that biomarkers, once validated, can be useful in identifying populations and individuals at risk in time to intervene effectively.