Three expert panels reviewed the 1991 Maryland School Performance Assessment Program (MSPAP) reading test. This was the first year of an assessment program designed to measure school progress toward, among other content areas, three reading outcomes: reading for literary experience, reading to become informed, and reading to perform a task. The MSPAP, given throughout the state, is a nontraditional, criterion-referenced performance assessment, which in 1991 required 9 hours of testing time over an 8-day period. The three panels, one consisting of experts with an instructional perspective, one with a curricular perspective, and one a psychometric perspective, independently addressed a variety of test quality issues after review of the test materials and a presentation by a test developer. Results indicate that a test such as the 1991 MSPAP is adequate to assess school progress in reading but may be confounded with writing and may not adequately measure progress in basic reading skills. Findings suggest that a test such as the MSPAP is useful for making curricular and instructional decisions, but that use of the test for making decisions about individual students was not supported. (Contains 24 references.) (Author)