This report represents the results of an investigation of the appropriateness of criterion-referenced tests (CRTS) for large-scale evaluations. First, the development and validation of CRTS, including the formulation and generation of CRT objectives, items, and score-interpretation schemes and dimensions of item and test quality, were examined to determine whether on theoretical grounds alone CRTS are suitable for large-scale evaluations. Second, the practical characteristics of CRTS were studied to determine if it is feasible to use currently available CRTS for large-scale evaluations. A set of criteria for selecting tests for evaluation purposes was devised and used to review 28 CRTS. A conclusion was reached that CRTS are not appropriate for use in large-scale evaluations for practical but not theoretical reasons. (Author)