A number of issues involved in the large-scale assessment of science and mathematics education were illustrated during the conduct of two national surveys: the 1977 National Survey of Science, Mathematics, and Social Studies Education and the 1985 National Survey, Science and Mathematics Education. Development of the survey questionnaires started with a series of questions from the sponsoring agency. However, explicit indicators of quality in science and mathematics education were not available; thus, the advisory committee developing the questionnaire had many debates over data collection needs. One fortunate decision of the 1977 study was to study the textbooks used. An unfortunate decision was to omit course enrollment statistics for different races and genders; this would have proved useful. The 1985 survey focused on getting information from principals and from teachers, since those respondents previously provided the most heavily used data. Acquiring approval for data collection was often a lengthy process. The 1985 survey emphasized teacher certification and subject matter qualifications. These items were difficult to analyze and still require further study. The 55-item 1985 Teacher Questionnaire and the 28-item 1985 Principal Questionnaire are appended. (GDC)