Work on the development of history performance assessments is described. So far, six complete sets of assessments in United States history have been developed as part of this project. Students are first assessed on their historical knowledge of the period through a short-answer test. They are then asked to write an essay explaining the positions of authors of the primary source texts provided. A series of studies conducted to determine how scoring rubrics should be developed found the best strategy to involve looking at differences between expert and novice performance. The use of task specifications controlled some of the variation commonly associated with performance tasks. Results from the research to date include: (1) development of a valid scoring scheme; (2) development of rater training procedures; (3) a task structure that reduces score variability; (4) distinguishing between assessment purposes and the utility of overall score and subscores; (5) detecting gender differences in scores; (6) supporting the validity of the measures through grade point averages and a scale ensuring effort through studies with 68 11th graders; and (7) addressing validity criteria with the same data. There are 27 tables and 4 graphs of study data, and a 40-item list of references. Five appendices provide samples of the content assessments and information about validity and reliability results. (SLD)