The traditional indicators for health status in population groups are based primarily on mortality rates and sections of disease receiving treatment. This has proved insufficient for planning and assessment of the efforts made by health systems to solve present day health problems. In order to develop better indicators, the concepts of health and disease must be analysed further and operational definitions must be established. As the choice of indicator is closely associated with the object of the investigation, it is necessary to define in every case the purpose for which the indicator is to be employed. This hold true also if attempts are made to build up a more general health index to illustrate important aspects of the conditions of health of a population. Many attempts of this type are in progress. Morbidity constitutes an important element in many attempts to assess the conditions of health of population groups. Measurement of disease is normally based upon clinical data, subjective data or behavioural data which each provide a picture of morbidity. The daily functional levels are employed in various ways as indicators of health. These may be fitness for work, mobility and social activity but in every case these have a close connection with current social and community conditions. Psychological problems assume an important place in our disease pattern but are extremely difficult to describe in population groups. The commencing point may be psychiatric disease in identified patients, extensive population investigations or indirect measurements which are considered to be related to psychiatric disease. All three methods tend to provide distorted pictures of the true situation. Various models have been employed in attempts to describe the general state of health. For example a health spectrum has been employed with axes for physical, mental and social health, which show definite mutual relationships or a great number of vectors have been built up to form a health profile. Many attempts have been made to establish one dimensional indices which may for example, be based upon functional level and expressions such as harmony and dissatisfaction. The danger involved in using methods such as these should not be overlooked. These methods imply simplification of the actual problems and they cannot replace more profound social medical analyses. There can be no doubt that better indicators are required for planning of health services and it is essential that efforts be made in Denmark to develop methods which are in agreement with our own needs and problems.