When the energy intake of (NZB X NZW)F1 female mice was reduced to 60% of the intake of simultaneously ad libitum-fed mice, the early death associated with autoimmune-based renal disease in this strain was greatly delayed. The length of prolongation of disease-free life depended not only on the decreased energy intake but also on the energy source. In the group of mice with 60% intake of a carbohydrate-free (i.e., high fat) diet, mean longevity was doubled as compared to that of ad libitum-fed mice. However, when the nonprotein energy was supplied by carbohydrate (sucrose and glycerol) the mean longevity was three times that of the ad libitum-fed groups, although survival times varied widely. With ad libitum feeding the nonprotein energy source did not significantly affect longevity. Clearly, although energy intake restriction provides significant influence on longevity, very high fat diets do not give the same protection as do high carbohydrate diets. The basis for this difference is not entirely clear and several explanations are possible.