A comparison was made of the effects of cholesterol feeding in normal rabbits and in rabbits rendered persistently diabetic by means of alloxan. In the two groups of animals hypercholesterolemia of comparable degree was induced by the feeding procedure. Nevertheless, the severity of the atherosclerosis of the aorta produced in the diabetic rabbits was much less than in the non-diabetic control animals. Indeed, a large proportion of the diabetic animals presented no atherosclerosis whatever. There was a similar inhibition of the deposit of lipid substances in the liver, spleen, and adrenal glands of the diabetic rabbits. The inhibition of the development of experimental cholesterol atherosclerosis which was associated with the presence of alloxan diabetes was independent of the administration of alloxan per se. It was not dependent on the sex or weight of the animal, nor upon the daily dosage of cholesterol, the form in which it was administered, nor the duration of cholesterol feeding. It was also independent of changes in body weight occurring during the course of our experiments and of the actual degree of hypercholesterolemia induced by the administration of cholesterol. In addition, there was no gross or histological evidence of a morphological basis for the inhibitory effect either in the aorta or in the other organs in which it was observed. Only two factors were observed to be consistently associated with the inhibition of the expected morphological effects of cholesterol feeding, namely, the diabetic state and a degree of visible lipemia considerably greater than that observed in the control animals. The inhibitory effect observed in these experiments would appear to depend upon some as yet undetermined factor or factors implicit in the diabetic state or closely associated with it. The experimental data presented demonstrate clearly that hypercholesterolemia is not the sole factor concerned in the genesis of experimental cholesterol atherosclerosis, but that another factor, or factors, rendered inoperative in our experiments must be essential to the production of the arterial lesions. In view of the inhibitory effect on the development of experimental cholesterol atherosclerosis observed in alloxan-diabetic rabbits, the effect of alloxan diabetes on the retrogression of such arterial lesions was studied in another series of experiments. No effect on retrogression could be demonstrated within periods lasting up to a maximum of 4 months after the cessation of cholesterol feeding. The results of our two series of experiments, considered together, indicate that the process of deposition of lipids in the arterial walls is governed by factors different from those that are operative in the process of removal of lipids after they have been deposited. The inhibition of the development of experimental cholesterol atherosclerosis in alloxan-diabetic rabbits must depend on interference with the process of deposition of lipids and not on a process of removal of lipids as fast as they are deposited. Our experimental results find no direct application to the problem of arterial disease in human diabetes. Nevertheless, the experimental procedures that we have employed provide a new basis for the design of further experiments directed toward the elucidation of the nature of the unknown factors that govern the process of lipid deposition in the walls of arteries.