In much of the developing world, if one wishes to know about the health status of people one needs to go to their homes and watch, interview, or examine them. Of course, persons who have come to hospitals or health centres are routinely described in the many data forms completed by the attending physicians, midwives, and other health workers. Unfortunately, for several reasons this information is of limited use in determining the health status of a defined population. First, sick persons may not seek the assistance of a trained health worker. Thus, using only provider data, the impact of illness in a given region would be underestimated. Secondly, the population being served by either local or regional medical care providers may not be easily identified. Thus, with too large or too small a denominator, the local disease rate would be either under- or over-estimated. Thirdly, the providers of care may be too busy to complete and submit the many forms required by most government administrators. Their omissions would result in fewer reported cases so that the local disease rates would be underestimated. Administrators and planners are well aware of these problems. As a result, they often question the value of the presented data for comparing one region with another so that the information has little effect on the decisionmaking process. Yet if 'health for all' is to become a reality, objective information on the needs of the community must be quickly gathered and evaluated so that resources can be allocated in the most efficient manner.