Reaction of a mammal to a given noxious agent is an inheritable property and there are known substantial differences even among particular strains of the same species. One example is the existence of different strains of mice especially raised to be particularly (and uniformly) sensitive to insults by certain chemicals. Another example is the existence of certain tumours that are common with some animals and unknown in others. Such is the case with thymic lymphoma: very common in certain mouse strains, but apparently with no real analogue in humans. Reverse situations undoubtedly exist. The inference is that the predictive value for human hazard varies for different animal experiments. Although the discovery of any toxic effect points to the need for close medical supervision of exposed persons in order to detect similar changes in man at an early stage, accurate quantitative extrapolation to man of the results of animal experiments is not feasible.