Biochemical techniques have been applied for years in clinical medicine to unravel disease indicators and mechanisms. The study design, resembling that of a case-referent study, has been to compare diseased and healthy individuals. Conclusive results have been obtained for diseases in which relatively few factors cause high risks. The same study design has been maintained with sophisticated methods of molecular biology, which have revealed many types of deoxyribonucleic acid lesions associated with cancer. It is likely that the relatively simple cross-sectional and case-referent studies require input from epidemiologists to unravel multifactorial etiologies. Longitudinal approaches may be the only alternative with which to evaluate the role and usefulness of the application of molecular biology markers in studies of cancer.