This paper presents a new method for health problem assessment, called IMP. It is based upon the premise that the critical step in the development of policy decisions isthe definition of health problems at the community level. The method is an extension of the Delphi technique but has been significantly improved by the determination of the critical variables in the problem set and by the incroporation of a structured logic system and of professional judgment. The study compares the IMP method with data gathered by traditional, epidemiological prevalence surveys and also with data from a computer simulation. The conclusion is unavoidable: the IMP method is an accurate as the other methods, at far less cost. The problem of preventable bureaucratic blindness has been used as an illustration of the application of this technique.