The prevalence of diabetes is predicted to increase dramatically over the next 30 years, mostly as a result of Type 2 diabetes. To manage this increased burden, the following measures must be implemented: diabetes care must be delivered in the most suitable locations; people with diabetes must remain in contact with the appropriate healthcare professionals; and health professionals must use the most effective methods to treat their patients. The structure for diabetes care suggested by the World Health Organization uses terminology that is familiar universally. The reality, in terms of what can be delivered at primary-and secondary-care level will, however, differ greatly according to the extent of service development and the resources available. The optimum forum for diabetes care should have the advantages of the primary-care setting and, when required, the main advantages of the secondary-care setting, i.e. specialized knowledge and more sophisticated facilities. In most healthcare systems, particularly those without financial disincentives to the patient to be followed up in hospital, the tendency remains for secondary care to 'capture' and retain patients for longer than is required clinically. The end results are the overloading of hospital services, inappropriate use of resources and the tendency for patients to default from care because of the relative inaccessibility and remoteness of the hospital service. Issues with special reference to different models of diabetes care, use of patient registers and clinical guidelines, and the perceptions of patients regarding the successful use of services are presented.