Genetic variations have generally been accepted to be the major mechanism for the differences observed in susceptibility to cancer. However, extensive investigations on genotype-disease associations have not produced consistent results. The inconsistency may be caused by many factors, such as improper study design, insufficient sample size, complexity of the traits under investigation, heterogeneity of the study subjects, incorrect assumptions about the underlying genetic architecture, misclassification of the disease, improper selection of potential alleles, and overinterpretation of the data. Besides these "traditional" factors, a recent problem is that the function of many variant genotypes is unknown, especially with regard to polymorphic DNA repair genes. Therefore, in addition to the genotype-disease relationship, it may be prudent to step back and investigate the fundamental role of genetic variation in the development of cancer. The latter studies may focus on understanding the genotype-exposure interactions and evaluating genotype-health risk associations. This review will emphasize the use of relevant biomarkers and polymorphic DNA repair genes for investigations. These data will be useful for a better understanding of the complexity of disease causation and development, developing new models for human disease, and identifying pathways for prevention of cancer.