Burgeoning levels of diabetes are a major concern for dialysis services, as diabetes is now the most common cause of end-stage renal disease in most developed nations. With the rapid rise in diabetes prevalence in developing countries, the burden of end stage renal failure due to diabetes is also expected to rise in such countries. Diabetic patients on dialysis have a high burden of morbidity and mortality, particularly from cardiovascular disease, and a higher societal and economic cost compared to non-diabetic subjects on dialysis. Tight glycaemic and blood pressure control in diabetic patients has an important impact in reducing risk of progression to end stage renal disease. The evidence for improving glycaemic control in patients on dialysis having an impact on mortality or morbidity is sparse. Indeed, many factors make improving glycaemic control in patients on dialysis very challenging, including therapeutic difficulties with hypoglycaemic agents, monitoring difficulties, dialysis strategies that exacerbate hyperglycaemia or hypoglycaemia, and possibly a degree of therapeutic nihilism or inertia on the part of clinical diabetologists and nephrologists. Standard drug therapy for hyperglycaemia (eg, metformin) is clearly not possible in patients on dialysis. Thus, sulphonylureas and insulin have been the mainstay of treatment. Newer therapies for hyperglycaemia, such as gliptins and glucagon-like peptide-1 analogues have become available, but until recently, renal failure has precluded their use. Newer gliptins, however, are now licensed for use in ‘severe renal failure’, although they have yet to be trialled in dialysis patients. Diabetic patients on dialysis have special needs, as they have a much greater burden of complications (cardiac, retinal and foot). They may be best managed in a multidisciplinary diabetic–renal clinic setting, using the skills of diabetologists, nephrologists, clinical nurse specialists in nephrology and diabetes, along with dietitians and podiatrists.