The introduction of the trial state assessment program into the design of the 1990 National Assessment of Educational Progress (NAEP) raises questions about differences across states in sampling and administration practices. In addition, questions about the general approach to comparing state data to national data need discussion. Subject areas covered by the NAEP include reading, mathematics, and science; state programs will only cover mathematics. NAEP samples are selected from both age and grade populations; 9-, 13-, and 17-year-olds and fourth-, eighth-, and twelfth-grade students will be involved. State assessments will only involve eighth-graders. The NAEP will cover private as well as public schools, while the state component will cover only the latter. NAEP will involve a deeply stratified, multi-level sampling plan, with oversampling of minority students and private schools. The NAEP sample will allow regional, but not state-by-state, comparisons. The target for the state-level program includes 2,000 eighth-graders selected from about 100 schools in each state. While special staff are provided for the NAEP, state programs will be conducted by local school staff under rigid guidelines. Special subsampling procedures have been developed for both national and state populations. (TJH)