This paper summarizes the standardized procedures and psychometric standards states must follow in developing performance-based assessment. The National Reporting System (NRS), which serves as the accountability system for federally funded adult education and literacy programs, requires that states have their local programs assess educational gain by selecting the assessments either published standardized tests or performance-based assessments most appropriate for their state. Performance-based assessments obtain richer information about students than do selected-response assessments, or standardized tests, and are often seen as more authentic because they resemble-real world tasks. But they cover less content, make it hard to generalize from one test to another, and are difficult and costly to score. In order to ensure that test scores resulting from performance-based assessments are valid and reliable according to the Standards for Educational and Psychological Testing, there are three steps in development. They are the following: (1) develop the tasks the test-takers will perform; (2) develop the scoring rubrics that will measure how well the test-takers performed on the tasks; and (3) pilot and field-test the assessment and scoring to determine the validity and reliability of the test items and scoring. Includes 4 references. (MO)