Diabetes mellitus (DM) is the most common cause of end-stage kidney disease and a major risk of morbidity and mortality. It is not clear whether medical management of DM has any significant beneficial effect on clinical outcomes at the end-stage of diabetic nephropathy with full-blown micro- and macro-angiopathic complications. Both loss of kidney function and dialysis treatment interfere with glucose homeostasis and confound glycemic control. Given the unique nature of uremic milieu and dialysis therapy related alterations, there have been some debates about reliance on the conventional measures of glycemic control, in particular the clinical relevance of hemoglobin A1c and its recommended target range of <7 % in diabetic dialysis patients. Moreover, a so-called burnt-out diabetes phenomenon has been described, in that many diabetic dialysis patients experience frequent hypoglycemic episodes prompting cessation of their anti-diabetic therapies transiently or even permanently. By reviewing the recent literature we argue that the use of A1c for management of diabetic dialysis patients should be encouraged if appropriate target ranges specific for these patients (e.g. 6 to 8 %) are used. We also argue that "burnt-out diabetes" is a true biologic phenomenon and highly prevalent in dialysis patients with established history and end-stage diabetic nephropathy and explore the role of protein-energy wasting to this end. Similarly, the J- or U-shaped associations between A1c or blood glucose concentrations and mortality are likely biologically plausible phenomena that should be taken into consideration in the management of diabetic dialysis patients to avoid hypoglycemia and its fatal consequences in diabetic dialysis patients.