- To evaluate the effectiveness of the non-mydriatic digital camera for diabetic retinopathy (DR) screening. Secondary purposes of the study were to characterize diabetic patients being screened for the presence of DR and to calculate the sensitivity, specificity, and positive predictive value of the test. All 6,962 consecutive patients with type 2 diabetes undergoing non-mydriatic digital retinal photography between January 1, 2009 and June 30, 2010 in eight community health clinics in the south of the country were included. Comparison of a random sample of patients who underwent non-mydriatic retinal photography, and who were also examined by an ophthalmologist with pupil dilation was also performed. The average age of all patients was 64.2 years. A total of 5,960 cases (85.6 % of all photographs) were of adequate quality for the diagnosis. DR of any degree was found in 1,092 (18.3 %) patients. Normal fundus pictures were found in 49.4 % of patients. In 32.2 % of cases, non-DR pathologies were found. Among cases in which DR was found, 73.3 % (801 cases) had mild non-proliferative retinopathy (NPDR), 7.1 % (77 cases) had moderate NPDR, 6.8 % (74 cases) had proliferative retinopathy, and 12.8 % (140 cases) had diabetic macular edema. Older patients had more chance of having poor quality pictures (p < 0.001 between patients older and younger than 70 years). When non-mydriatic fundus photography was compared with dilated fundus examination by an ophthalmologist, sensitivity of 99.3 %, specificity of 88.3 %, and positive predictive value of 85.3 % were found. Non-mydriatic digital retinal photography is an efficient method for DR screening. The test has high sensitivity and specificity. The test, as performed in community health centers in the south of the country, contributed to the early diagnosis of >1,000 cases of DR. Many patients can be followed up in a fast and efficient way, although the test cannot replace a complete eye examination after pupil dilation mainly in older people. Other non-DR sight-threatening ocular pathologies can be also detected.