This paper presents a discussion of outcome assessment that puts into context how measurement has evolved to its present state. Several types of testing and assessment options are considered against a background of validity. Criterion-referenced measurement is discussed extensively in terms of history, field study, identity problems, intellectual conflict, social conflict, and test design. The conflict between criterion-referenced tests and norm-referenced tests, development of norm-referenced tests, domain-referenced testing, problems with domain referenced tests, new approaches to content specification, quality control, integration of testing and instruction, and the narrow definition of testing are discussed as subjects for concern. Following this, a special model of evaluation adapted to the problem of new technologies is detailed. In order for the model to assess new technologies as desired, the information must: (1) provide an enhanced documentary base for the processes of new technology development; (2) use state-of-the-art evaluation methodology, including both quantitative and qualitative approaches to measurement; and (3) provide policy feedback to the supporting agencies. Criterion referenced measurement is advocated for the assessment of instructional technology outcomes, with the caveat that such measurement is difficult. An 18 page bibliography is appended. (LMO)