Abstract Background and Aims Renal involvement in type 2 diabetes is mostly presumed to be due to diabetic nephropathy, however a significant majority of diabetic patients can have pure non diabetic kidney disease (NDKD) or NDKD superimposed on diabetic kidney disease (DKD). Kidney biopsy cannot be routinely performed for all diabetic patients and hence short of conclusive biomarkers we need to explore various factors that can predict the occurrence of NDKD. Method We conducted a retrospective review of all native kidney biopsies conducted in patients with type 2 diabetes at our institute to identify the prevalence and factors that predict NDKD. The demographic data, clinical data, laboratory parameters, and histological results of the patients were obtained from their medical records. Binary logistic regression analysis was performed to evaluate the predictive factors for NDKD. Patients were categorized into 3 groups-DKD alone (presence of Kimmelstiel-Wilson nodules, capsular drops, fibrin caps and absence of any other NDKD features), NDKD along (presence of vasculopathy, interstitial fibrosis and specific glomerular changes in absence of classical DKD changes) and lastly DKD+NDKD (presence of histological changes of both DKD+NDKD). Results We analyzed a total of 69 patients. The mean (SD) age of the cohort was 51.94 ± 12.7 years and males constituted the majority (68%). Around 2/3rd of the entire cohort had hypertension, with no significant differences among the groups. Mean 24-hour urinary proteinuria was 6.6 ± 4.9 g/TV and 15.9% patients had microscopic hematuria. The mean (SD) duration of diabetes was 6.9 ± 6.2 years. Patients with pure DKD, pure NDKD and NDKD superimposed on DKD constituted 46.3%, 33.3% and 20% respectively of the cohort. The most common indication for renal biopsy in our cohort was rapid GFR decline in 38 patients (54.2%), followed by massive proteinuria in 27 (38.5%) and lastly positive serological work up for multiple myeloma in 5 patients (7.1%). Around 2/3rd of pure DKD patients in our cohort had diabetic retinopathy (DR), whereas only around 1/5th (26%) of pure NDKD patients had DR (p=0.005). DR was absent in around 74% of our diabetic patients who had pure NDKD. Non diabetic kidney disease in our cohort was seen in 46.3% patients and combined with DKD in 20% of patients. Among patients with NDKD, the most common cause was membranous glomerulonephritis (MGN) (43.4%) followed by acute tubulointerstitial nephritis (ATIN) (17.3%). Among combined DKD and NDKD, most common histological diagnosis was MGN (32.4%) followed by ATIN (16.2%). Independent factors predicting NDKD were shorter duration of diabetes (OR 0.74, CI 0.59–0.94, P=0.01) and absence of diabetic retinopathy (OR 0.15, 95% CI 0.09–0.26, P = 0.01). Conclusion Judicious use of kidney biopsy revealed NDKD in nearly half of T2DM patients especially in those with short duration of diabetes and absence of diabetic retinopathy (DR). Kidney biopsy is strongly recommended for T2DM patients with atypical presentation and in absence of DR.