It has been recommended that the National Assessment of Educational Progress (NAEP) specify comprehensive exercise domains to measure academic achievement, and provide a national item pool to measure the objectives in these domains. These domain specifications and item pools would serve to satisfy the increasing demand for valid, accurate, and detailed testing and interpretation. Item response theory (IRT) can be used to tailor tests to different purposes and populations. The domain would consist of a set of objectives, hierarchically organized, spanning some range of cognitive learning outcomes. The corresponding item pool would include items written according to the specifications for each objective, appropriately reviewed, field tested, and calibrated. Nationally developed domains and item pools could be used to compare states' curriculum guidelines, describe achievement trends, and quantify achievement in a manner superior to traditional norm referenced tests. NAEP, at the Federal level, provides the natural vehicle for such efforts. Domain descriptions in each curriculum area should represent the content at various levels of aggregation and reflect skill hierarchy, referent generality, and instructional sequence. Implementation issues requiring further consideration involve comprehensiveness of outcome domains, including items for all objectives and affective objectives; choice of IRT model, standards and criteria; and reporting of results. (GDC)