Diabetes is common in the old, defined here as aged 80 plus; in Melton Mowbray 11–14% of octogenarians had diabetes1. What are the implications of the landmark United Kingdom Prospective Diabetes Study (UKPDS) for them? UKPDS showed that intensive glycaemic control with sulphonylurea or insulin reduced microvascular endpoints; there was an approximately 25% risk reduction of microvascular endpoints at l0 years reducing the risk from 11.4 to 8.6%2. Approximately half of these patients were aged over 65 at l0 years study duration and there were very broad inclusion criteria. Unfortunately, 50% of elderly diabetic subjects are dead within five years of census, 3 and 40% of subjects newly diagnosed by glucose tolerance test are also dead within 5 years. Several studies have now shown that improved glycaemic control leads to improved cognitive function4-7 and improved glycaemic control leads to improved quality of life, 4, 8, 9 although there is one contradictory study10. Thus as a geriatrician I still aim for glycaemic control as good as I can reasonably and safely achieve, but my justification is cognitive function and quality of life. The UKPDS Hypertension in Diabetes embedded study11 is of major interest since intensive blood pressure reduction significantly reduced many adverse endpoints such as heart failure, stroke (fatal and non-fatal), microvascular disease and diabetes related death; the Kaplan-Meier plots appear to show that many of these differences are apparent at 4 to 5 years. These findings are particularly relevant to the old diabetic person not only because of the relatively short time to derive benefit, but also because these are the co-morbid conditions that affect the British diabetic retiree. Dornan's comparison of elderly/old subjects known and not known to have diabetes12 showed that the diabetic person was significantly more likely to have had a previous stroke, cognitive impairment, congestive cardiac failure, neuropathy and blind registration. Other studies have similarly shown that the elderly diabetic person is at increased risk of dementia, 13 which is presumably predominantly vascular in origin14. As the authors of UKPDS pointed out, one should also look at the Systolic Hypertension in the Elderly Program (SHEP) study:15 elderly subjects with systolic blood pressure greater than 160 mm Hg were randomised to active treatment with a thiazide diuretic, or to placebo; this study had 650 octogenarians (8% diabetic) and a 5 year follow up. There are subgroup analyses of diabetic subjects16 and old subjects, 7 but not old diabetic subjects. In SHEP, treatment of octogenarians reduced more strokes than treatment of younger subjects15.17 The absolute benefit of antihypertensive treatment was greater in the diabetic than in the non-diabetic group;16 in the diabetic group active treatment significantly reduced major cardiovascular events, and all coronary heart disease events. However, SHEP accepted only about 1% of subjects referred and excluded subjects receiving insulin, so one wonders how generalisable are the results. The Systolic Hypertension in Europe Trial (Syst-Eur)18 had 441 octogenarians (24% diabetic);18 treatment with a long acting calcium channel blocking agent aimed to reduce systolic BP to<150 mm Hg. The study was terminated after a median follow-up of 2 years because a significant result had been obtained. Subgroup analysis of diabetic subjects19 shows a greater absolute benefit from treatment than in the non-diabetic subjects; in the diabetic group active treatment significantly reduced cardiovascular death, all cardiovascular events and strokes. Subgroup analysis of old subjects17 reveals a non-significant decrease in strokes. Treatment also reduced the incidence of dementia20. Diabetic patients were included in Syst-Eur if their diabetes was controlled, i.e. insulin treatment was not an exclusion criterion. There are two worries regarding antihypertensive treatment in the old: although it decreases strokes, there is a non-significant increase in mortality;17 however, given the greater absolute benefit of treating diabetic subjects, this may be even less applicable to the diabetic octogenarian. Secondly, the subjects in these trials may be fitter than the patient in your clinic. Given the comorbidity of the old diabetic person, a study of diabetes in frail old folk is required, but unlikely to happen. However, HYVET is an ongoing placebo controlled trial of antihypertensive therapy in octogenarians; its inclusion criteria are broad and its pilot study recruited 1284 subjects so that one anticipates a large diabetic subgroup; other outcomes of interest such as falls, quality of life and confusion will be assessed. treat as per SHEP or Syst-Eur targets (e.g. systolic BP <150 mm Hg); conclude that there is limited data on treating run of the mill octogenarians, and await new data or enter the patient into HYVET. At present, I cautiously treat as per SHEP and Syst-Eur but I am very wary of inducing postural hypotension. Thus the importance of UKPDS is to direct our attention to the old diabetic subject's blood pressure, which will hopefully improve disability free survival.