The human body is endowed with a large number of xenobiotic chemical metabolizing enzymes, a significant proportion of which are polymorphic and thus render one individual at greater or lesser risk than another of chemically-induced disease. All examples of genetic polymorphism of chemical metabolizing enzymes have been reviewed in relation to their potential to activate and detoxicate procarcinogens and promutagens. Many examples are cited whereby phenotype can act as a carcinogenic risk factor. With the availability of a large amount of DNA sequence data for chemical metabolizing enzymes there has emerged a number of polymerase chain reaction (PCR) strategies aimed at discerning one metabolic phenotype or another. This is seen as a very positive and democratic scientific development, widening the franchise for studies of disease risk. Nevertheless, it is argued that, at these early stages with many laboratory-based scientists scarcely familiar with epidemiological study design, a cautious approach should obtain when interpreting single studies.