A striking familial aggregation, high concordance in monozygotic twins, and segregation analyses all support the notion of a major genetic susceptibility for type 2 diabetes. Several genetic loci that result in autosomal dominant insulin secretion defects and onset of type 2 diabetes before 25 years of age have been identified. These discoveries have uncovered a new pathway, the transcription factors hepatocyte nuclear factor 1α and 4α, and possibly other genes in these regulatory pathways. However, these genes play at most a small role in the more typical late-onset type 2 diabetes. Studies of candidate genes for insulin secretion and insulin sensitivity have likewise explained little about diabetes susceptibility. Common amino acid variants of the insulin receptor substrate 1 gene appear to have a small role, as does a common variant of the intestinal fatty acid binding protein and the β3-adrenergic receptor gene, but inconsistent findings argue that these polymorphisms do not explain most cases of diabetes susceptibility. Likewise, an unexplained association of the sulfonylurea receptor gene with type 2 diabetes seems unlikely to represent a major locus. Genome-wide scans remain in their infancy, but several promising findings may lead to new genes. These include the NIDDM1 locus on chromosome 2q, evidence for a locus somewhere on chromosome 20, and evidence in both Pima Indians and a US population of Northern European descent for a locus near the apolipoprotein A-2 locus on chromosome 1. Future studies, particularly combined analyses of groups currently working independently, will better define new diabetes genes.