The many significant changes that have come about in the mental health service of different countries in the last two decades have made administrative and political decision making in the distribution of care services a much more hazardous business. In the United States, for example, the number of occupied psychiatric beds fell from 550, 000 in 1955 to 190, 000 in 1977. England and Wales experienced similar if less pronounced changes, while in the Federal Republic of Germany the same trend became apparent some five years ago, although here the initial hospitalisation rates were lower. Enquiry into the real needs for various forms of mental health care, especially the need for hospital beds, for places in homes and hostels, and for specialist out-patient treatment, has now become a funda­ mental aim of research in social psychiatry. To achieve this goal by epi­ demiological concepts and methods, including the investigation of true morbidity rates and the estimation of the related need for care, must be used.