Background and aim Most studies show that diabetes increases with migration and urbanization. Previous studies from Greenland have shown inconsistent associations between cardiovascular risk and urbanization. Thus, the aim was to study the association between diabetes and urbanization among Greenland Inuit. Methods A total of 3089 adult Inuit aged 18years and older participated in a geographically representative, population-based study 'Inuit Health in Transition Study'. The examination included a 75g oral glucose tolerance test and anthropometric measurements. Information on socio-demographic characteristic and health behaviour was obtained by interview or questionnaire. The participants were categorized according to degree of urbanization into three groups based on current place of residence: (1) participants living in towns (>2000 inhabitants), (2) participants living in small towns (<2000 inhabitants) and (3) participants living in villages (<500 inhabitants). Results The total prevalence of diabetes was 9% of which 79% were previously unknown. Nine per cent had impaired glucose tolerance and 19% had impaired fasting glycaemia (IFG). Compared with towns, odds rations (ORs) for diabetes and impaired fasting glycaemia were higher in small towns [OR diabetes=1.5 (1.0-2.3), OR IFG=1.9 (1.2-2.3)] and villages [OR diabetes=1.2 (0.8-1.9), OR IFG=1.3 (0.9-2.0)], whereas no association was seen for impaired glucose tolerance. The inverse association between urbanization and diabetes and impaired fasting glycaemia persisted after adjustment for relevant confounders. Conclusion Diabetes and impaired fasting glycaemia decreased with urbanization contrary to the results of most studies. It appears that Greenland Inuit follow the pattern usually observed in industrialized countries with the highest risk of diabetes in the lower socio-economic groups.