IN the early stages of planning a sample survey one is ordinarily concerned with two different sets of questions, both relating to the survey's feasibility. First of all one is concerned with the possibility of obtaining the required information from the type of respondent considered. Thus, if the subject under investigation is the prevalence rate of certain ailments in the general population, one class of questions relates to whether such information would or could be supplied by individuals or by doctors, whether a suitable questionnaire could be devised, and the like. A second class of questions relates to economic feasibility. Assuming that a class of respondent can-be found who can supply accurate information of the type required, does one have sufficient resources-funds, personnel, time-to survey a sufficient number of respondents to yield results of desired precision? To a considerable extent, questions of economic feasibility are answered if one knows how large a sample of respondents is required. In consequence an initial approximate determination of sample size is one of the early steps in the planning of a survey. We propose to illustrate some of the factors involved by consideration of a specific survey. As is well known, the Public Health