In a previous study we demonstrated the presence and diagnostic usefulness of circulating rhodopsin mRNA in the assessment of diabetic retinopathy (DR). In the present study we investigated three further retina-specific markers in blood to determine their suitability as markers of DR. The markers were RPE65, retinoschisin, and melanopsin. Whole blood was collected from diabetic patients and healthy controls into PAXgene Blood RNA tubes and RNA was extracted using the PAXgene Blood RNA System. Quantitative real-time PCR was used to quantify mRNA for RPE65, retinoschisin, and melanopsin. beta-actin mRNA was used for normalization. RPE65, retinoschisin, and beta-actin mRNA were detected in 100% of subjects; melanopsin was not detected in either controls or diabetic patients. Circulating RPE65 mRNA concentration was 63% higher in diabetic patients than in healthy individuals (P= 0.019), whereas retinoschisin showed no change between the two groups. Compared with healthy controls, circulating RPE65 mRNA concentration was higher in diabetics with no retinopathy (30%; P= NS), background DR (93%; P= 0.01), preproliferative DR (20%; P= NS), and proliferative DR (107%; P= 0.004). Compared with diabetics with no retinopathy, levels of RPE65 mRNA were also significantly higher (60%) in the presence of proliferative DR (P= 0.029). In contrast, levels of retinoschisin mRNA were lower in background DR (34%; P= 0.033), preproliferative DR (43%; P= 0.026), and proliferative DR (47%; P= 0.038) compared to that in diabetics without retinopathy. We conclude that not all retina-specific mRNA species are detectable in circulation (e.g., melanopsin). This may be related to differences in expression levels for the individual markers. Both RPE65 and retinoschisin were detectable and demonstrated contrasting trends in diabetics with and without retinopathy. In combination with rhodopsin, RPE65, and retinoschisin, mRNA may offer a useful tool in developing a blood test for DR.