Normal vision depends on the normal function of retinal neurons, so vision loss in diabetes must ultimately be explained in terms of altered neuronal function. However, to date relatively little attention has been paid to the impact of diabetes on the neural retina. Instead, the focus of most research has been primarily on retinal vascular changes, with the assumption that they cause altered neuronal function and consequently vision loss. An increasing body of evidence suggests that alterations in neuronal function and viability may contribute to the pathogenic mechanisms of diabetic retinopathy beginning shortly after the onset of diabetes. This view arises from neurophysiological, psychometric, histopathological and biochemical observations in humans and experimental animals. The collective evidence from past and recent studies supports the hypothesis that neurodegeneration, together with functional changes in the vasculature, is an important component of diabetic retinopathy. The authors invite other investigators to include the neural retina as a component of their studies so that the pathogenesis of diabetic retinopathy can be understood more clearly.