Out of level testing is defined and reasons for administering out of level tests, especially in compensatory education programs are explained: (1) to achieve optimum test reliability; (2) to test student achievement--provided out of level test scores are interpreted in terms of the appropriate in-level population; (3) to avoid carelessness and inattentiveness or frustration and guessing caused by students' taking tests at an inappropriate level; and (4) to avoid large numbers of students scoring at chance level. Two procedures which allow out of level tests to be converted to in-level scores are described using publisher produced raw score-to-percentile conversion tables, or standard score-to-percentile conversion tables. Chance level scores, names of expanded standard score scales, and worksheets to convert out of level scores are appended. (MH)