The discussion centers around the amount and kind of participation that should be asked of a student taking reading and oral comprehension tests. For each test there are four levels of participation; zero, limited, extended, and complete. The test designer should create a situation which will encourage the subject to give maximum expression, because his language in the test is a representative sample of what he can produce in a real situation. This also enables the corrector to give a more accurate evaluation of the student's language ability. The most effective oral test includes an interview and a discussion section to permit the most extensive use of the language. Various ways of evaluating linguistic structures are suggested, with emphasis on organization of test goals. (FB)