Lung cancer is now the major cause of cancer deaths in the USA and is an increasingly significant cancer worldwide. Biological markers could be used to prevent lung cancer by allowing more timely and precise understanding of the role of environmental factors. So far, biological markers that can serve as carcinogen dosimeters have been investigated in only a small number of pilot studies of populations with current or past exposure to lung carcinogens. We describe several of our collaborative studies involving smokers, various worker populations, lung cancer cases and controls in order to illustrate the advantages and the limitations of 'molecular epidemiology'. The enzyme-linked immunosorbent assay (ELISA) with antibodies to polycyclic aromatic hydrocarbon (PAH)-DNA adducts has been used in conjunction with one or more of the following: physicochemical techniques to monitor carcinogen adducts on haemoglobin, cytogenetic methods to quantify sister chromatid exchange (SCE) and chromosomal aberrations, and Southern and western blot analyses of oncogene activation. Increased levels of markers of biologically effective doses have generally been seen in exposed and high-risk groups when compared to controls. We have also observed significant background levels of such markers and interindividual variation in levels of certain biological markers resulting from exposures to carcinogens. Thus, these methods may be particularly useful in identifying segments of the population that have received a significant effective dose and hence can be considered to be at elevated risk of cancer. Such studies are necessary to validate laboratory methods and lay the groundwork for more definitive molecular epidemiological investigations of lung cancer.