Despite the use of a wide range of different methodologies and diagnostic criteria, epidemiological studies have shown that Type 2 (non-insulin-dependent) diabetes has a global distribution and its prevalence varies from country to country, in different ethnic groups in the same country, and between the same ethnic group undergoing internal or external migration. Rural-urban and migration studies indicate that change towards a 'Westernized' lifestyle is associated with a dramatic increase in the prevalence rates for Type 2 diabetes. Between populations, comparisons are confounded by a wide range of different ascertainment rates, survey methodologies, and diagnostic criteria for diabetes. However, low prevalence rates for Type 2 diabetes are seen in Eskimos and populations of the Far East, while the highest are seen in American Indians, urbanized Pacific Island populations, and migrant Asian Indians. Available evidence suggests that these latter groups have a genetic susceptibility to Type 2 diabetes ('diabetes genotype') and that the disease is unmasked by environmental factors. There appears to be a spectrum of interaction between genetic and environmental factors--in certain populations the genetic role may be more important than environmental, or vice versa. Epidemiological studies, apart from their value in population screenings and case-finding, have contributed to the new classification and diagnostic criteria for diabetes and our understanding of risk factors and host characteristics in the aetiology of Type 2 diabetes.