This commentary reviews the recent reports of the DIRECT study. 2 This study has explored the role of the Ang II receptor blocker, candesartan as a prevention or a treatment strategy in type 1 and type 2 diabetes subjects with, or at risk of, diabetic retinopathy. The findings from these trials extend our understanding of the role of medical treatments in diabetic retinopathy. This condition is currently managed when the retinal lesion is already advanced with essentially destructive, albeit useful, treatments such as laser photocoagulation. Diabetic retinopathy remains the major cause of blindness in the working age population although the prevalence of vision-threatening retinopathy appears to be decreasing and in general progression rates have been on the decline. The underlying explanation for the changes in the ‘natural history’ is not clear but may be as a result of improvement management of key risk factors such as glycaemic control and blood pressure, based on the results of studies such as the DCCT and UKPDS. 4 Indeed, as recently confirmed in the follow-up of the Steno-2 trial, multi-factorial intervention including blood pressure, glucose and lipid control does reduce the development and progression of retinopathy. The rationale for the DIRECT study was partly based on a range of experimental studies that had first identified a retinal renin-angiotensin system and, second, demonstrated, using both ACE inhibitors and Ang II antagonists including candesartan, functional and structural effects on various retinal disorders including diabetic retinopathy as well as a model of retinal neovascularisation, retinopathy of prematurity. 7 Furthermore, with increasing evidence that certain growth factors implicated in enhanced retinal permeability to proteins and angiogenesis such as VEGF are Ang II dependent, it was considered that interruption of the reninangiotensin system would have direct effects on pathological processes implicated in diabetic retinopathy. The DIRECT study was essentially three separate placebo controlled trials, recruiting a total of over 5,000 subjects with type 1 and type 2 diabetes. It included an arm known as Prevent 1, which studied the role of candesartan in preventing or delaying the onset of retinopathy in normotensive type 1 diabetes subjects. Other arms included Protect 1 and Protect 2, which examined the role of the Ang II antagonist in retarding or reversing progression of retinopathy in type 1 and type 2 diabetes subjects, respectively, with evidence of early retinopathy. 2 It should be noted that only in a subgroup of subjects in the Protect 2 arm who had concomitant hypertension could additional antihypertensive drugs be added. This multi-national study with over 300 sites worldwide, involving a follow-up of subjects for four years, demonstrated some benefits of candesartan treatment in the various studies. However, some effects observed were borderline and of debatable clinical significance. With respect to Prevent 1, there was an 18% reduction in two-step progression of retinopathy of borderline statistical significance. However, although only a post-hoc analysis, there was a 35% reduction in three-step progression of retinopathy, a parameter that has been previously considered the ‘gold standard’ of diabetic retinopathy studies. The findings from Protect 1 were not particularly impressive with only modest benefits of doubtful clinical significance observed. However, in Protect 2, a 34% increase in regression of retinopathy was observed. With this complex series of findings it is difficult to see how these results would translate to clinical practice. Without achieving the pre-specified primary endpoint, it may be difficult, based on this trial, for candesartan to be registered by major regulatory agencies as a retino-protective agent. Nevertheless, with increasing use of agents which interrupt the renin-angiotensin system in diabetic individuals for other indications such as hypertension and nephropathy, this study provides the clinician with a lower threshold for commencing such agents in the diabetic population. Furthermore, the positive findings from the DIRECT study build on previous, much smaller, less well designed studies such as the EUCLID study, which demonstrated clear effects on retinopathy progression with the ACE inhibitor lisinopril. The issue of blocking the renin-angiotensin system in type 1 diabetic subjects has been further explored in the RASS study which reported in a smaller number of subjects a reduction in the development and progression of diabetic retinopathy with either the ACE inhibitor, enalapril or the Ang II antagonist, losartan. This additional study, although performed in a relatively small cohort of type 1 diabetic subjects builds on the underlying postulate that the renin-angiotensin system plays a key role in the diabetic retina and that retinal protection may be afforded by various different approaches to interrupt the renin-angiotensin system.