The vocational education program needs of the nation's large cities were assessed by a mail survey of vocational educators in 106 major urban centers. Several procedural problems in generating appropriate instrument content and initial problems in obtaining data from respondents were encountered. Goal and objective statements generated through a review of current literature and the opinions of a sample of experts were not acceptable to local vocational educators. To overcome these problems, an open-ended survey and content analysis procedure was used to generate a set of goals and objectives judged acceptable by local vocational educators. Early exploration of alternative strategies indicated that a discrepancy methodology would reduce the rate of survey returns. Consequently, the study relied on a methodology requesting respondents to rate the importance of educational needs. Because of the large number of goals and objectives identified by the first step of the procedure, multiple forms were developed and utilized in the study. A response rate of over 65 percent was obtained. This experience in conducting a national needs assessment study causes us to question the applicability of the discrepancy approach for large studies. Alternative strategies of direct ratings of needs may be more feasible. (Author)