Traditionally, occurrence of microvascular disease (retinopathy) has been used to define the glucose threshold by which diabetes should be defined. But these studies have come from mostly white/European/middle eastern populations. Do all ethnic groups have similar thresholds for diagnosis? Are certain diagnostic tests better for making the diagnosis of diabetes than others in select race/ethnic groups? The pathophysiology of type 2 diabetes may differ and we need to define subtypes of T2DM.