A favourable prognosis for a penetrating corneal graft is largely dependent upon the integrity of the endothelium of the donor cornea, so that any method which will allow the condition of this delicate cell layer to be evaluated in potential transplant material will assist in promoting success.At present three such methods are in limited use: (i) The histochemical method of Robbins, Capella, and Kaufman (I965) and Kaufman, Capella, and Robbins (I966) employs staining with para nitro blue tetrazolium to demonstrate the viability of corneal cells. The procedure kills all cells, and it is therefore necessary to assume that the condition of the corneal endothelium is the same in both eyes of a donor pair at any given time. The endothelium of one cornea is evaluated, and if satisfactory, the graft material is taken from the other.(2) The permeability test (Stocker, King, Lucas, and Georgiade, 1966, I970) uses the permeability of the endothelial cells to the vital dye trypan blue as an index of cell viability. The method evaluates the endothelium of the cornea actually to be used as the graft, but subjects it to much risk of damage during the manipulation involved.(3) Direct visual examination of the corneal endothelium through the rest of the cornea of intact donor eyes with high power specular-reflection microscopy (Hoefle, Maurice, and Sibley, 1970) is the most simple and least traumatic method, but it does not necessarily follow that an endothelium which appears 'normal' morphologically is functionally competent, and the converse is true. This paper reports the results of animal experiments on another method of evaluating donor corneal endothelium, but now from the point of view of its net functional efficiency in maintaining stromal deturgescence, a major requirement of most corneal grafts. The endothelium of the actual transplant material is evaluated, in the intact eye, with no risk of damage to the cells. Concurrently the thickness of swollen corneae is reduced closer to normal, and the morphology of the endothelium can be examined by the technique of Hoefle and others (I970).The method depends upon the temperature reversal phenomenon of the cornea (Davson, I 955;Harris and Nordquist, I955). When a fresh normal eye is refrigerated to below io'C. in a moist environment, as in an orthodox eye bank, its cornea swells, and when rewarmed to physiological temperature the cornea thins down again to normal values. This reversal in corneal thickness in relation to temperature is due to great retardation of the corneal metabolic rate in the cold, resulting in the cessation of the outward going active fluid transport system which is located in the endothelium (Mishima and Kudo, I967; Maurice,