BACKGROUND: Diabetic retinopathy is a leading cause of visual loss. Hypothesis-generating data from cardiovascular outcome trials suggest that fenofibrate therapy may reduce the progression of diabetic retinopathy. OBJECTIVE: To determine whether treatment with fenofibrate reduces the progression of diabetic retinopathy. DESIGN AND METHODS: We conducted a parallel-group, double-masked, placebo-controlled clinical trial of fenofibrate. A web-based algorithm allocated participants to treatment arms by minimisation. SETTING AND PARTICIPANTS: The trial was positioned within NHS Scotland's Diabetic Eye Screening Programme. Adults with diabetes and non-referable retinopathy or maculopathy (based on Diabetic Eye Screening retinal image grading) were eligible. INTERVENTIONS: Study treatment was mailed to participants' homes. Participants who were eligible at the screening assessment entered an active pre-randomisation run-in during which they took 145 mg fenofibrate. After randomisation, participants received 145 mg fenofibrate tablets or placebo. Study treatment was taken daily in those with normal renal function, or on alternate days in those with impaired renal function. MAIN OUTCOME MEASURES: The primary outcome was a composite of developing referable diabetic retinopathy or maculopathy, or requiring treatment for diabetic retinopathy or maculopathy. Incremental cost-effectiveness was assessed in terms of the primary outcome and per modelled quality-adjusted life-year gained. DATA SOURCES: Data were obtained from 6-monthly interviews by research nurses and linkage to national healthcare data sets. Selected adverse events were adjudicated by study clinicians masked to treatment allocation. RESULTS: One thousand four hundred and eighty-four participants entered the pre-randomisation run-in, of whom 1151 were randomised. The primary outcome occurred in 131 (22.7%) of 576 participants assigned fenofibrate and 168 (29.2%) of 575 participants assigned placebo (hazard ratio 0.73; 95% confidence interval 0.58 to 0.91; p = 0.006) over a median of 4.0 years. Any progression of retinopathy or maculopathy, and development of macular oedema were also reduced. There was no effect on visual function, quality of life, or visual acuity. Fenofibrate use resulted in a non-significant reduction in 6-monthly health service costs (mean difference -£101, 95% confidence interval -£243 to £42), leading to dominance over standard care and a high probability of cost-effectiveness. Based on modelling (assuming no difference in background healthcare costs by treatment allocation), fenofibrate led to a small increase (£6) in cost for a small gain (0.02) in quality-adjusted life-years; incremental cost-effectiveness ratio £406 per quality-adjusted life-year gained. The probability of cost-effectiveness was 79-86% at thresholds of £20,000-30,000 per quality-adjusted life-year gained. LIMITATIONS: Early Treatment Diabetic Retinopathy Study retinopathy grading is considered the gold standard, but it is not used in large-scale retinal screening programmes; Diabetic Eye Screening grading is based on Early Treatment Diabetic Retinopathy Study but is less granular. CONCLUSIONS: Fenofibrate was clinically effective and cost-effective for reducing the progression of diabetic retinopathy compared with placebo among participants with early retinal changes. FUTURE WORK: LENS participants will be followed for 10 years to assess the long-term effects of fenofibrate therapy. FUNDING: This synopsis presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number 14/49/84.