Direct epidemiological observations suggest that exposure to high levels of urban air pollution may result in increased risk of lung cancer, sufficient to account for a few (approximately 1-3) percent of total lung cancer incidence. Extrapolation from occupational exposure and risk data suggests that among potential carcinogens present in polluted urban air, polycyclic aromatic hydrocarbons (PAHs) may make a major contribution to air pollution-associated lung cancer risks. The use of biomarkers of genotoxocity in large-scale population studies may help to reduce the uncertainty involved in the assessment of such risks, especially those associated with relatively low pollution levels such as nowadays found in many Western cities. Increases in biomarkers of exposure to urban air PAHs as well as biomarkers of early effects have been detected in situations of relatively high levels of air pollution (e. g., ambient PAH concentrations of the order of a few tens of micrograms per cubic meter). Evidence has also been found about the modulation genetic damage accumulation in different individuals by polymorphisms in genes involved in the activation or detoxification of PAHs, especially of polymorphisms GSTM1 and CYP1A1 genes. However, the inconsistencies in the currently reported effects of genetic polymorphisms suggest that additional factors may also be important in the modulation of individual susceptibility to the accumulation of PAH-derived genetic damage. Biomarkers studies in populations exposed to relatively low ambient PAH concentrations (below 20 microg/m(3)) have not demonstrated clear dose-related effects (e.g., on DNA adduct levels), possibly because of the existence of multiple sources and routes of human exposure to PAHs in addition to inhalation of urban air (including, for example, home heating, environmental tobacco smoke and diet), and the consequent difficulty of adequately and specifically assessing atmospheric air-related exposure. This makes it imperative that molecular epidemiology studies be designed in such a way as to allow adequate assessment of exposure to urban air PAHs at the individual level and over short-, medium- and long-term time periods which correspond to the expression times of different biomarkers.