Theoretical reasons and experimental evidence indicate that a no-effect level generally cannot be expected for genotoxic carcinogens; as a consequence, in quantitative risk assessment the capability of distinguishing genotoxic from non-genotoxic carcinogens is of fundamental importance in order to identify relevant levels of human exposure. According to generally accepted guidelines, the standard three-test battery for the detection of genotoxic compounds consists of: (i) an in vitro test for gene mutation in bacteria; (ii) an in vitro test in mammalian cells with cytogenetic evaluation of chromosomal damage and/or a test that detects gene mutations; (iii) an in vivo test for chromosomal damage using rodent hematopoietic cells. This test battery is designed to avoid the risk of false negative results for compounds with genotoxic potential, but it cannot be taken for granted that the risk is completely eliminated. As a matter of fact there are some chemicals, classified by the International Agency for Research on Cancer (IARC) as probably or possibly carcinogenic to humans, which gave consistent negative results in this test battery, and in contrast provided positive results in other not routinely employed genotoxicity assays. The failure of the standard test battery in detecting some genotoxic carcinogens is attributable to several causes, but the principal of them are the following ones: in vitro, the artificial metabolic activity of the liver S9-mix, and the different biotransformation of chemicals in cells of different type and from different animal species; in vivo, the pharmacokinetic behaviour of the test compound, and its possible species-, sex- and tissue-specificity.