Introduction: Early and accurate detection of onset of diabetic nephropathy is crucial for a timely intervention and for improving outcomes in the management of diabetic nephropathy. There is a need of new biomarkers to predict the onset of nephropathy earlier than the conventional markers. The aim of this study was to measure the association of urinary periostin level in type 2 subjects and healthy controls and to evaluate its predictive value to diagnose renal injury. Methods: This was a cross sectional study in which urine samples were collected from 50 healthy controls and 360 type 2 diabetic patients. The diabetic cases were further grouped as; normoalbuminuria (n=95), microalbuminuria (n=130) and macroalbuminuria (n=135). The urinary levels of periostin was measured by using an enzyme-linked immunosorbent assay kit. Observations: Urinary periostin levels were found to be significantly elevated in all the diabetic patients compared to the controls. In multivariate regression analysis increased urine periostin level significantly correlated with aging, high albuminuria and decline of eGFR. For differentiating normalbuminuria, microalbumiuria and macroalbuminuria the best selected cut-off values with optimum sensitivity and specificity were >2.11, 2.74 and 7.81 respectively with predicted sensitivity and specificity of 88% and 100%, 98.1% and 66.7% and 100% and 69.9% respectively. Therefore, urinaryperiostin demonstrated moderate to high sensitivity and specificity for diagnosing diabetic nephropathy. Conclusion: Increased levels of urinaryperiostin significantly correlated with aging, high albuminuria, declinement of eGFR and can be detected in the urine of the patients with type 2 diabetes before the onset of significant albuminuria. Urinary periostin excretion may be used as an early biomarker of renal injury.