The concept of a threshold of activity of a genotoxic agent is primarily based upon considerations of protective mechanisms and multiple cellular targets, which require inactivation before a toxic response is produced. In this paper, we have considered and evaluated the influences of compound metabolism, DNA lesion formation, mutation induction and sequence content, aneuploidy induction and the influence of repair enzymes upon genetic endpoints produced by both DNA reactive chemicals and by those chemicals which modify non-DNA cellular targets. Thresholds of activity have been evaluated by critical analysis of the published literature and original data analysing both the role of sequence context upon point mutation induction and DNA repair mechanisms upon the sensitivity of cultured cells to the induction of aneuploidy. In the case of DNA reactive chemicals, the presence of a threshold of chemical activity will be dependent upon cellular activities such as those of the Phase II enzymes reducing the activity of chemicals before lesion formation takes place and/or those of the DNA repair enzymes which reduce the proportion of DNA lesions which are processed into DNA sequence changes. Under such conditions, a given exposure of a DNA reactive chemical does not produce a linear or semi-linear increase in DNA lesions or in mutation frequency. However, even when these protective mechanisms are overwhelmed by the high exposures of genotoxic chemicals the biological effects of a genotoxin may be influenced by the sequence context of the gene under consideration. Here, we demonstrate that point mutations are detected at relatively higher frequencies in the non-coding introns compared with the coding exons. Many of the base changes detected in the exons do not produce amino acid changes in the proteins coded for by the genes being monitored for mutation induction. Both sequence context and the types of base changes induced may provide a "buffering" effect reducing the biological consequences of mutation induction. Spindle damaging chemicals, such as colcemid and vinblastine, induce aneuploidy by modifying the numbers of spindle fibres which regulate the segregation of chromosomes during mitosis and meiosis. The redundancy of spindle fibres in the dividing mammalian cell leads to the prediction that only chemical exposures which damage most, if not all, of the fibres will lead to the induction of polyploidy and/or aneuploidy. Such predicted thresholds of chemical activity can be observed when both chromosome loss and non-disjunction are measured in wild type cultures. However, we observed a substantial increase in sensitivity to aneugenic chemicals when measurements were made in primary cell cultures derived from xerodoma pigmentosum and trichothiodystrophy patients. Further studies are necessary to evaluate the consequences of the genetic background of tester strains upon the nature of the dose-response curve of aneugenic chemicals.