The author traces reasons to support his contention that the state of the art in assessing learning disabled students is not good. Among issues examined are the following: use of tests for purposes other than those for which they were intended; technical adequacy of currently used tests (standardization, reliability, validity); the use of deficit scores to identify the learning disabled; bias in assessment; bias following assessment; and assessment of abilities vs. assessment of skills. The questions of who is to test and who is to make decisions regarding placement are addressed. Six factors, including increased disillusionment on the part of classroom teachers with the kinds and quality of assessment and increased emphasis on the rights of children are seen to be bringing about change. The importance of differentiating assessment, viewing tests as samples of behavior, and using technically adequate tests is underscored. (CL)