A challenging goal of molecular epidemiology is to identify an individual's risk of cancer. Molecular epidemiology integrates molecular biology, in vitro and in vivo laboratory models, biochemistry and epidemiology to infer individual cancer risk. Molecular dosimetry of carcinogen exposure is an important facet of molecular epidemiology and cancer risk assessment. Carcinogen macromolecular adduct levels, cytogenetic alterations and somatic cell mutations can be measured to determine the biologically effective doses of carcinogens. Molecular epidemiology also explores host cancer susceptibilities, such as carcinogen metabolism, DNA repair, and epigenetic and genetic alterations in tumor suppressor genes. p53 is a prototype tumor suppressor gene and is well suited for analysis of mutational spectrum in human cancer. The analyses of germ line and somatic mutation spectra of the p53 tumor suppressor gene provide important clues for cancer risk assessment in molecular epidemiology. For example, characteristic p53 mutation spectra have been associated with: dietary aflatoxin B1 exposure and hepatocellular carcinoma; sunlight exposure and skin carcinoma; and cigarette smoking and lung cancer. The mutation spectrum also reveals those p53 mutants that provide cells with a selective clonal expansion advantage during the multistep process of carcinogenesis. The p53 gene encodes a multifunctional protein involved in the cellular response to stress including DNA damage and hypoxia. Certain p53 mutants lose tumor suppressor activity and gain oncogenic activity, which is one explanation for the commonality of p53 mutations in human cancer. Molecular epidemiological results can be evaluated for causation by inference of the Bradford-Hill criteria, i.e., strength of association (consistency, specificity and temporality) and biological plausibility, which utilizes the "weight of the evidence principle."