Genotoxic chemicals can damage the genetic material of humans as well as that of organisms living in the environment. With respect to adverse effects, alterations induced in the germ line, leading to alterations in the genetic make-up of populations, are of primary concern in ecosystems, because somatic changes, even if they lead to a loss of individuals, will not be critical in populations with a large reproductive surplus. This is different in human toxicology where genetic alterations in germ cells as well as in somatic cells of any individual are of concern. Increased frequencies of mutations and related genetic alterations in the gene pools of individual species or populations in ecosystems have to be judged against the background of spontaneous mutations that have enabled species to survive and adapt in changing environments since the beginning of life on our planet, and which have played an important role as the substrate for evolutionary developments. Examples of the selection of altered phenotypes (and genotypes) in response to environmental pollution and environmental stress are melanism in moth populations, metal resistance in plants, insecticide resistance in insects and malaria resistance in humans. Pollution, in general, can represent a stress factor selectively leading to a change in genetic make-up. In addition, environmental genotoxins can directly alter gene pools. A change in the genetic constitution may be advantageous for certain populations living in stressful conditions, but may present a disadvantage for others, including man. Examples are (i) the induction of (pesticide) resistance, (ii) the increased virulence of pathogens, (iii) alterations of host ranges of pathogenic forms or the appearance of new virus types and (iv) subtle changes in parasite-host or predator-prey relationships. Basically the release of genotoxins into the environment should be avoided because massive exposures may affect the reproductive capacity of many species, and modest exposures may lead to an enhanced instability of ecosystems and may provoke specific adaptations to stressful situations. Furthermore, the uncontrolled presence of genotoxins in any compartment of the natural environment is an unwanted situation, in particular also from a human point of view. In addition we need novel quantitative approaches in order to make quantitative risk estimates possible.