Type 2 diabetes is one of the fastest growing public health problems worldwide. Both environmental (e.g. physical activity, obesity, and diet) and genetic factors are involved in the development of type 2 diabetes. The associations between physical activity and diabetes risk have been assessed by a number of prospective studies and clinical trials. The results from these studies consistently indicate that the regular physical activity during occupation, commuting, leisure time or daily life reduces the risk of type 2 diabetes by 15-60%; and lifestyle intervention, including counselling for physical activity, nutrition, and body weight, can reduce the risk of type 2 diabetes by 40-60% among adults with impaired glucose tolerance and by about 20% among general individuals. In the past decade, studies using traditional linkage analysis and candidate-gene association approach have found dozens of genes harboring common variants that were related to the common-form type 2 diabetes. However, most reported associations are lack of reproducibility, except TCF7L2, PPARG, CAPN10, and KCNJ11. Since 2007, seven genome-wide association (GWA) studies emerged to generate a list of new diabetes genes. The genetic effects are largely of moderate size. These findings provide novel insight into the diabetes etiology and pave new avenue for predicting the disease risk using genetic information. In addition, data especially those from intervention trials display preliminary but promising evidence that the genetic variants might interact with physical activity in predisposing to type 2 diabetes. The gene-environment interactions merit extensive exploration in large, prospective studies.