A discussion is presented of the view that whole-body irradiation of an animal can, in certain circumstances, speed up natural aging. In survival studies of chronically irradiated CBA mice, kept under certain conditions so that external disturbing factors were minimized, mean life shortening of a group of mice is roughly proportional to the mean accumulated radiation dose. Up to st run. Concent 15 rem/day, there was no reduction in dispersion of survival times even though the mean survival time was nearly halved; the cumulative mortality curves were merely displaced from each other, not contracted; the Gompertz plots were roughly of equal slope. The duration of development was unaffected by radiation and was statistically the same for irradiated and control animals; a value of about 180 days was deduced. The linearity of the dose response is consistent with a cumulative injury due to irreversible changes in cells. The linearity argues against the primary process being one of chromosome structural damage and suggests, rather, somatic gene mutation. The fact that short single doses of x or gamma radiation are st run. Concent 3 times as efficient as the same doses delivered chronically at low rates indicates a doserate factor that is notmore » expected for gene mutation. Moreover, the large size of the doubling dose (thousands of rems) is surprising for a gene mutation unless cell selection for the radiation mutants is much more severe than for the natural mutants. The site of the change may not be in the chromosomes, but a cytoplasmic phenomenon. According to the simplest theory, the life shortening per rem is numerically equal to the ratio of the rate of cell change produced by unit dose rate to the spontaneous rate. (TCO)« less