For any given level of exposure to a carcinogen, only a proportion of exposed individuals will develop cancer. Interindividual differences in susceptibility at some stage of the carcinogenic process must be postulated. One contributing factor is variation in the activity of metabolizing enzymes responsible for conversion of procarcinogens to proximate carcinogens. There is also a wide spectrum of DNA repair capability within the general population. At one end are the genetic instability syndromes characterized by extreme sensitivity to carcinogenic exposures and high rates of cancer in homozygotes of these traits. Less extreme differences in mutagen sensitivity can be demonstrated by a quantitative assay of chromosome breaks induced by in vitro mutagen exposure. Two case-control studies of patients with previously untreated upper aerodigestive tract cancers have demonstrated mutagen sensitivity to be an independent risk factor for the disease after controlling for the effects of tobacco and alcohol. Mutagen sensitivity also may have prognostic relevance. There was a fourfold elevated risk of developing multiple primary cancers in mutagen-sensitive patients. There are also data suggestive of familial aggregation of cancer in first-degree relatives of mutagen-sensitive patients (twofold risk for having one first-degree relative with cancer and sixfold risk for having two or more relatives with cancer). The preventive implications of identifying markers of carcinogen sensitivity are manifold.