Ageing is not adaptive since it reduces reproductive potential, and the argument that it evolved to provide offspring with living space is hard to sustain for most species. An alternative theory is based on the recognition that the force of natural selection declines with age, since in most environments individuals die from predation, disease or starvation. Ageing could therefore be the combined result of late-expressed deleterious genes which are beyond the reach of effective negative selection. However, this argument is circular, since the concept of 'late expression' itself implies the prior existence of adult age-related physiological processes. Organisms that do not age are essentially in a steady state in which chronologically young and old individuals are physiologically the same. In this situation the synthesis of macromolecules must be sufficiently accurate to prevent error feedback and the development of lethal 'error catastrophes'. This involves the expenditure of energy, which is required for both kinetic proof-reading and other accuracy promoting devices. It may be selectively advantageous for higher organisms to adopt an energy saving strategy of reduced accuracy in somatic cells to accelerate development and reproduction, but the consequence will be eventual deterioration and death. This 'disposable soma' theory of the evolution of ageing also proposes that a high level of accuracy is maintained in immortal germ line cells, or alternatively, that any defective germ cells are eliminated. The evolution of an increase in longevity in mammals may be due to a concomitant reduction in the rates of growth and reproduction and an increase in the accuracy of synthesis of macromolecules. The theory can be tested by measuring accuracy in germ line and somatic cells and also by comparing somatic cells from mammals with different longevities.