Type 2 diabetes is rapidly increasing worldwide thereby posing a severe burden on individual and public health. Although this increase can be ascribed to both genetic and environmental factors, until 2007, the genetic factors have been largely restricted to a few candidate genes based upon our limited knowledge of the pathogenesis of the disease. In an attempt to identify underlying genetic loci, several genome-wide association studies (GWASs) for type 2 diabetes have been completed during the last few years. GWASs have proven to be an unbiased research tool to identify genetic loci that play an important role in many metabolic diseases and thus successfully contributed to our understanding of the disease pathogenesis. However, in type 2 diabetes, like in most polygenic diseases, variations in or near known genes today explain only about 10% of the inherited risk of the disease. Undoubtedly not all of the biological pathways that, when altered, can lead to disease are fully understood. Thus, the initial wave of GWAS and a limited follow-up represent only the beginning of systematic studies of gene discovery as a tool for dissecting the complexity of type 2 diabetes. The overall aim of this thesis was to test the ability of genetic variants to predict future type 2 diabetes and to explore the mechanisms by which they increase the risk of the disease. First, in the large prospective studies from Sweden and Finland, we did not only show that genetic risk factors modestly improved the prediction of future type 2 diabetes as compared to clinical risk factors alone, but we also demonstrated that the value of the genetic risk factors increased the earlier they were measured during life. Carriers of high genetic risk for type 2 diabetes showed impaired capacity of β-cells to increase insulin secretion in response to obesity and insulin resistance. Effects of genetic variants associated with type 2 diabetes and/or glucose and insulin levels were not only restricted to β-cells, but also included α-cells and the crosstalk between the gut and the islets, the enteroinsular axis. A variant in the most important Asian type 2 diabetes gene, KCNQ1, resulted in impaired insulin secretion and also increased risk of future type 2 diabetes in Scandinavians. A variant in the PCSK2 gene is associated with reduced insulin secretion and increased risk of type 2 diabetes, adding yet another potential type 2 diabetes susceptibility gene to the steadily increasing list. Taken together, the comprehensive knowledge on how genetic factors influence the risk of type 2 diabetes and identifying novel pathways in the pathogenesis of the disease most likely will pave ways to new therapeutic strategies. Though genetic tests to predict the disease cannot be offered yet, it may be possible to use them on a population level to reduce the number of individuals needed to be included in intervention studies aiming at prevention of type 2 diabetes. In conclusion, studies presented in this thesis provide an important contribution to the diabetes research and to our understanding of the pathogenesis of type 2 diabetes.