Diabetic retinopathy (DR), is a major cause of blindness, caused by prolonged diabetes. However, this morbidity is largely preventable and treatable. Therefore identification of early changes in DR is important to help management of disease. The progression of DR from prolonged diabetes involves complex metabolic de-regulations. ¹H NMR-based metabolomics is one of the promising technique to assess metabolic changes in bio-fluids. To our knowledge this is the first ¹H NMR metabolomics study of serum in DR patients. The aim of the present study was to explore dysregulation in metabolites expression for the differential serum metabolic profile of DR patients using ¹H NMR–based metabolomics analysis. Venous blood samples from 6 DR patients with type 1 diabetes (diabetes duration, 12.2 ± 4.5 years; HbA1c, 8.4% ± 1.4%) and from 6 age-matched non-diabetic volunteers were collected. Samples were incubated at room temperature for 45 min to allow clotting followed by centrifugation at 1500 × g for 10 min. 200μL of the supernatant (serum) were transferred into sterile cryovials, lyophilized and stored at − 80°C until further use. For NMR experiments, lyophilized serum samples were weighed and dissolved in D₂O. 600μL of sample was transferred to 5 mm NMR tubes. 1D spectra were acquired using 700MHz Agilent Spectrometer. The parameters used for 1D NMR spectra using CPMG sequence were: ns, 64; relax delay, 14; pulse angle, 90°; reference, 0.5 mM TSP. NMR data were processed in MATLAB using NMRLab/MetaboLab (1). Post-processing steps included scaling, alignment, exclusion of water signal and TSP signal, binning at 0.006 ppm. Statistical analysis included principal component analysis (PCA) and PLS- DA multivariate analysis using PLS toolbox in MATLAB.Figure 1 shows the representative NMR spectrum (expanded region 3.2 ppm to 3.9 ppm) obtained from controls (red) and diabetes patients (blue). Principal component analysis (PCA) analysis showed as score plot in Figure 2 depicts separation between two categories of samples. Serum samples of DR patients showed a trend towards lower or negative values of PC1. To get more detailed insight into the data, PLS-DA was also used. Figure 3 shows PLS-DA revealing more clear separation between the two groups.¹H NMR based metabolomics analysis of serum samples provides an insight into the metabolic fingerprints of DR. The results of the present study revealed a vital association of metabolic dysfunction with DR. The PCA and PLS-DA analysis could distinguish metabolic profiles of DR patients from healthy controls in spite of limited number of samples. It should be noted that blood serum represents the effects of metabolism contributions from different organs and metabolic processes of body. The next step of the study would be to identify the different specific metabolites and their pathways that are altered in DR compared to healthy controls. Using such a study, we may observe the severity of disease on the basis of metabolic fingerprints of serum. This, in turn, may serve as a platform not only for future pathophysiological studies but also for the screening of molecular targets for a more efficient therapeutic intervention of this disturbing complication of diabetes.