Abstract Background and Aims Diabetes (DM) is considered one of the major causes of progressive CKD even when it does not induce proteinuria. However, the condition of diabetes without proteinuria is not universally considered to be accompanied by a reduced survival of the patient. The aim of the study was to verify whether non-proteinuric diabetic CKD patients presented a different risk of mortality compared to non-diabetic and non-proteinuric CKD patients. Method Two matched cohorts of CKD patients were obtained from the PIRP database, that contains clinical and laboratory data from about 30,000 patients who have been followed since 2005 in the Emilia-Romagna Region (Italy). In the period 2005-2017, 2651 subjects with DM2 in CKD-EPI stages 3a to 5 and without evidence of proteinuria that entered the PIRP register surveillance were matched with as many non-diabetics patients without proteinuria by gender, age class, CKD stage. To assess the presence of a differential mortality risk in diabetic patients, a multivariable Cox regression model was fit, adjusted for several potential clinical confounders, using multiple imputation to estimate missing data and taking into account patients’ clusterization into nephrology units to correct standard error estimates. Results In the 13 years examined, deaths were more frequent in patients with DM2 (38.7% vs 30.7%, p<0.001) as well as CV events before and after PIRP enrolment (52.2% vs. 42.9%, p<0.001), while dialysis initiation rate was lower (Fig.1). Furthermore, in diabetics there was a higher percentage of overweight or obese patients (38.9% vs 19.7%, p<0.001), lower levels of hemoglobin (12.3 g/dl vs 12.7 g/dl, p<0.001) and higher values of triglycerides 163 mg/dl vs 132 mg/dl, p<0.001). 10-year survival was significantly lower in diabetics (22.9% vs 36.4%, log-rank test: p <0.001, Fig.1). Multiple Cox regression identified a higher mortality risk for diabetics (HR = 1.326, p<0.001) after adjusting for sex, age, baseline eGFR, hemoglobin level, BMI, smoking status, presence of tumors or cardio-vascular comorbidities and cohort effect (5-years period indicator). Conclusion In patients with CKD and non-proteinuric nephropathy, concurrent diabetes was associated to a higher frequency of CV events and represented an independent risk factor for overall mortality. Thus, in diabetic patients, even in absence of proteinuria, we have to intensify therapeutic strategies to control the so-called CV intermediate risk factors (tobacco use, low-density lipoprotein cholesterol, hypertension, hypercoagulable states, obesity, etc.)