The hitherto published studies of HLA antigens in diabetic retinopathy are reviewed. The existence of HLA associated genes conferring susceptibility (or resistance) to the development of diabetic retinopathy has neither been proved nor ruled out. Possible explanations of the lack of unanimous conclusions are discussed. Further studies in caucasians and other ethnic groups are warranted: future studies should be internationally accepted, well-defined diagnostic criteria and classifications of diabetes and diabetic retinopathy, and study and control groups should be carefully matched for other genetic and nongenetic risk factors. Future studies should include HLA D/DR typing.