A needs assessment of the National Assessment of Educational Progress (NAEP) is presented. It deals with cost, design and technical issues, and utility. Suggestions include cost reduction via assessment schedule cutbacks and re-use of released NAEP exercises; and a shift from federal to private funding by selling NAEP exercises with interpretative materials to schools or individuals. A major unresolved question concerns the validity of inferences which can be drawn from the aggregated results. NAEP validation procedures constitute content validation. However, content validation does not necessarily constitute validity evidence at all, if validity evidence must bear on the interpretations that are warranted on the basis of test or assessment results. This means that more work needs to be done on the construct validation of the NAEP results as they are now commonly interpreted, i.e., with results aggregated across exercises. NAEP needs to become more clear in reporting aggregate results by specifying the facets of variance over which it is, and is not, attempting to generalize. Two strategies are suggested for the utility of NAEP results: developing norms for NAEP exercises, and making NAEP exercises and data more readily available to independent investigators. (PN)