The carcinogenic potential of heterocyclic amines (HCAs) in humans has yet to be established. Epidemiologic studies of colon cancer using crude surrogates for HCAs exposure (e.g., doneness of meat) have produced inconsistent results. To improve exposure assessment of HCAs, we have developed a database of HCA concentrations in commonly consumed meat items cooked by various techniques and degrees of doneness. HCA type and level are dependent on multiple factors, including type of meat (e.g., steak, chicken, bacon), cooking technique (with substantial variability present even within high temperature cooking methods), place of preparation (e.g., home, restaurant, or 'fast-food' restaurant), as well as the degree of doneness and surface browning/charring. We have developed a questionnaire with meat photographs linked to this database, which we are using in a variety of case-control and cohort studies of cancer etiology. In addition, we have carried out a metabolic study of HCA exposure among 66 subjects to identify biomarkers of HCA exposure which may be useful in epidemiologic studies. These studies should help clarify the role of HCAs in human carcinogenesis, and eventually allow an estimation of the cancer burden in the population attributable to these compounds.