The hypothesis that enhanced knowledge of assessment rubrics by teachers and thus by students results in improved student achievement was studied in the context of the development of mandatory high school assessments for the Maryland State Department of Education. Rubrics were under development to score constructed-response items in the content areas of English, biology, mathematics, and government. Each of the state's 24 local education agencies was represented for 2 or 3 content areas. Forty-six pairs of teachers provided data. Half (the experimental group), received training in the state's Core Learning Goals, instructional strategies, and the use of rubrics as instructional tools. The remainder (control group), did not receive training. Data analyzed were effect sizes by form by item type within teacher pair. Results provide some empirical support for instructional uses of rubrics. Support was strongest for constructed-response items in biology, but was also seen for selected-response and constructed-response items in algebra. Neither positive nor negative effects were seen for English or government. Whether positive effects remain at the same levels across different types of rubrics remains for additional study. An appendix contains the generic rubrics. (Contains two tables and eight references.) (SLD)