Type 2 diabetes mellitus is amultifactorial disease with an obvious polygenetic component. Using family-based linkage analysis, candidate gene-based genotype-phenotype association analysis, and the by far most successful method of hypothesis-free genome-wide association analysis, a total of 403 independent, mostly common gene variants in 243 gene loci significantly associated with type 2 diabetes risk have been hitherto identified. The gene variant with the most robust effect on type 2 diabetes risk is located in the TCF7L2 (transcription factor 7-like 2) gene locus and reduces incretin sensitivity of pancreatic -cells and limits the therapeutic response to drugs that make use of the incretin axis. Such pathomechanistic insights are currently only available for very few of the type 2 diabetes genes. Pathomechanistic investigations are hampered by gene-gene, gene-lifestyle, and gene-environment interactions. This is one of the main reasons why the impact of the risk alleles on type 2 diabetes prediction is still very limited.