Attempts to appraise the present state of knowledge on the accuracy of consumer surveys in describing markets for hypothetical or proposed goods and services. The paper is primarily concerned with those types of survey questions that were being developed by RAND to evaluate the social impact of the programs of the National Heart and Lung Institute. Relevant social science literature was reviewed and interviews were conducted with persons both inside and outside RAND who were familiar with survey techniques. The questions posed dealt with reducing chances of death or disability with a medical treatment; this type of question required that the respondent place himself in a hypothetical situation and state his likely action, given a set of circumstances. No literature was found reporting on surveys of individuals for the introduction of new medical products or procedures. Nor was much literature found that examines the correspondence between what people say they will do and the actions they subsequently take. 28 pp. Bibliog.