Objective: Alteration of renal haemodynamics is a hallmark of the disease with decreased bioavailability of nitric oxide (NO) being one of the most prominent disturbances. NO synthesis can be blocked by L-arginine analogs monomethyl-L-arginine (L-NMMA), asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA). Additionally, other effects of ADMA have been reported such as up-regulation of intra-renal RAS or toxic effect on endothelial progenitor cells. Therefore metabolism of un-/methylated arginines appears to be important pathogenic pathway in the progression of diabetic nephropathy (DN). The aim of the study was to assess the overall balance of the NO regulating metabolites in subjects with various degree of DN and their potential predictive value. Design: cross-sectional analysis of T1DM and T2DM diabetics with variable degree of DN with subsequent follow-up of the non-ESRD cohort for median 39 [IQR 21-58] months. Setting: diabetics regularly followed at the Diabetes and Nephrology Units of University Hospitals Brno and living in the South Moravia region, Czech Republic. Patients: 438 diabetics with following inception DN staging: normoalbuminuria (n=80), persistent microalbuminuria (n=127), persistent proteinuria (n=174) and ESRD (n=57, reg. hemodialysis). CKD stages were defined by measured GFR (creatinine clearance). Main outcome measurements: plasma ADMA, SDMA and L-arginine measured using HPLC with fluorescence detection, calculated ratio of plasma L-arginine/ADMA, a total of 6 tagging SNPs in the DDAH1 and DDAH2 genes , clinical end-points: (1) renal (progression of DN by stage or reaching ESRD stage), (2) major cardiovascular event (non-fatal myocardial infarction or stroke, amputation), (3) renal or cardiovascular mortality (i.e. complication of ESRD, fatal myocardial infarction or stroke, sudden death) and (4) all-cause mortality. Results: Plasma ADMA, SDMA and L-arginine differed between the four stages of DN with SDMA gradually increasing from normoalbuminuria to ESRD (solely renal excretion) while L-arginine and ADMA did not express such a pattern (the latter metabolised by DDAH1 and 2). We found significant negative correlation of ADMA and SDMA with GFR , while L-arginine did not correlate. No genetic were ascertained for ADMA and DDAH1 and 2 variants after the correction for multiple comparisons. Using Kaplan-Meier time-to-event analysis we identified signif. differences in renal end-point between SDMA tertile groups and in all-cause mortality between ADMA and SDMA tertile groups . The predictive value of SDMA still lasts when analysing subset of subjects with mild to moderate CKD at the baseline only , P=0.0015. Constructed tertiles of L-arginine/ADMA ratio exhibited signif. biological gradientin cardiovascular and all-cause mortality, the lowest tertile being associated with the lowest survival rate. Conclusions: (1) plasma SDMA but nor ADMA appears to reflect renal function in subjects with various degree of DN, however in case of ADMA enzyme degradation clear genotype (DDAH1 and 2) x phenotype relationship was not ascertained. SDMA is a predictor of DN progression in patients with none to moderate CKD. Furthermore (2) un-/methylated arginine status is related to cardiovascular and all-cause mortality in our cohort of patients, in particular not isolated ADMA levels but L-arginine/ADMA ratio serves as a predictor of adverse outcome.