In 1982-83, Riverside Unified School District (California) selected an achievement test to provide both norm- and criterion-referenced (dual objective) data for its achievement testing program. The test selection process was focused around eight evaluation aspects: prescreening of tests, content matching, instructional evaluation, processing evaluation, psychometric evaluation, input from interested staff and parents, cost analysis, and empirical preparation. Despite considerable effort to blend curricular and technical considerations, only the compromise of functional level testing was able to avoid a direct conflict between the curriculum committee's preference for the Comprehensive Tests of Basic Skills (CTBS) content and format, and the technical committee's concern for its ceiling effects. The CTBS covers 25-71 percent of the curriculum, though it does not assess about 40 percent of the curriculum judged measurable by a paper and pencil test. Although it is possible to wring content-referenced information from a standardized achievement test, the experience in Riverside suggests that the content analysis needs to be thorough and that interpretation of test results relating to the curriculum should be restricted to those specific areas in which sufficient match and coverage have been verified. (BW)