Background and aims: Recent analyses have suggested to stratify type 2 diabetes (T2D) into distinct subgroups characterized by different disease progression, risk of complications and frequencies of genetic variants. Applying cluster analysis, we aimed to evaluate the distribution of SNPs associated to T2D risk in subjects belonging to the VNDS cohort. Methods: 836 drug-naïve, GADA-negative, newly diagnosed T2D subjects underwent hyperinsulinemic euglycemic clamp to assess insulin sensitivity (IS) and 5h-OGTT minimal modeling to estimate β-cell function (BF) expressed as derivative (DC) and proportional (PC) control. All subjects have been genotyped for 23 TagSNPs belonging to genes previously associated to BF/IS and were stratified in several subgroups data-driven cluster analysis of age, HbA1c, BMI, BF, IS. Results: At diagnosis, 4 clusters have been identified: severe insulin resistance diabetes (SIRD; 22.1%), severe insulin-deficient diabetes (SIDD; 12.5%), mild age-related diabetes (MARD; 47.5%) and mild obese diabetes (MOD; 17.9%). The SIRD group, characterized by higher BMI (36.6±5.5 kg/m2) and lower IS (382.5±17.4 µmol/min/m2 BSA) (both p<0.001), was enriched by minor alleles of 3 SNPs involved in BF and abnormal fat distribution (rs483109-G, rs13387347-A within G6PC2 and rs17236708-G of FTO), and by the major alleles of SNPs within TCF7L2 (rs7901695-A, rs790314-G) and PPARG (rs13099643-A), belonging to BF and IS pathways, respectively. The MARD group, known to have T2D late onset (64.7±5.5 years), showed high frequency of the minor allele A of rs558994 within CACNA1E and major allele C of rs17832252 in GCK and major allele G of rs11196205 within TCF7L2, all genes involved in BF. The SIDD group showed the worst BF (DC:356.4±33.0 (pmol/m2 BSA)·(mmol·L−1·min−1)−1; PC: 34.9±2.5 [(pmol/min/m2 BSA)/(mmol/L)]) and the higher prevalence of the minor allele G of rs6544647 in THADA, known to be involved in BF. Finally, the MOD group, characterized by early T2D onset (45.4±6.4 years) and moderate obesity (BMI=30.8±4.0±kg/m2), was enriched with the minor alleles A of FTO rs11114831 and rs1477092, commonly related to obesity. Conclusion: In VNDS cohort, the phenotype and genotype-driven strategies allowed to identify distinct subgroups. This data might help to capture the heterogeneity of disease presentation, the response to treatment and risk of complications.