Epidemiologic evidence on the relation between ambient air pollution exposure and cancer is reviewed. The well-documented urban/rural difference in lung cancer incidence and the detection of known carcinogens in the atmosphere gave rise to the hypothesis that long-term exposure to air pollution may have an effect on lung cancer risk. However, problems inherent in assessing adequately the exposure of interest led to considerable difficulties in evaluating this effect. Routinely measured air pollutants do not include, as a rule, established carcinogens, and air pollution measurements usually come from fixed-site monitors, making it difficult to estimate individual exposures, especially long-term. The nature of the exposure and associated measurement problems made ecologic comparisons a natural way to approach the study of air pollution effects on lung cancer risk. The descriptive/ecologic studies which have been undertaken after 1950 often had problems with inadequate control of confounding, but, on the whole, provided evidence compatible with the hypothesis that urban and industrial air pollution may have an effect on lung cancer risk. The results of several case-control and cohort studies are described in the present review with emphasis on the exposure metric used. These studies, which control for important potential confounders, suggest that urban air pollution may be a risk factor for lung cancer, with estimated relative risks in the order of up to about 1.5 in most situations.