Two research in the cellular mechanisms of diabetic retinopathy have been of major recent interest—the pathogenesis of the earliest functional and anatomic lesions of the disease and the pathogenesis of the retinal neovascularization that occurs in its late stages. Of these two research problems, the first is probably the most perplexing. What are the earliest structural and functional abnormalities in the retinas of diabetic subjects? What is their cause? How do they lead to further abnormality and, ultimately, to the clinical picture of advanced diabetic retinopathy with loss of vision, with which we are all familiar? In the early 1960s, Cogan, Kuwabara, and their colleagues,1 in the pages of the Archives, described the results of their trypsin digest studies of the retinal vessels of human diabetic subjects. They described the earliest anatomic lesions that have been detected thus far in diabetic retinopathy, namely, loss of the intramural pericytes (or