We have created a database of published and unpublished studies on genetic susceptibility to environmental carcinogens, by bringing together many single studies that are too small to give definitive answers on the role of metabolic genes in cancer susceptibility. Possible participants were identified through a literature search of case control studies published up to June 1999 on metabolic gene polymorphisms and cancer, and invited to send their data sets without personal identifiers. Individual data from a total of over 30,000 subjects (52% cancer cases, 48% controls) have been collected. The most common type of cancer is lung, followed by bladder, head and neck, and breast. Demographic data, such as age, sex, and race were obtained for almost all the subjects. Main exposures, such as smoking, alcohol, occupational exposure were also included in a portion of the data set. The simultaneous presence of two gene polymorphisms has been assessed in 3535 controls and 3445 cases. An Advisory Committee has evaluated and approved 8 proposals to analyze the available data. This project allows the study of the main effects of genes, gene-exposure effects, ethnic and geographic differences in allele frequencies, gene-environment and gene-gene interaction as possible risk factors for cancer.