A study was conducted to compare five different techniques for determining the best means of assessing a reader's background knowledge and for discovering if it is possible to elicit different amounts of information from the same subject by simply varying techniques. Subjects were 32 third through sixth grade students, divided into groups of skilled and less skilled readers. The five assessment techniques used were (1) a free recall task about snakes; (2) 55 structured probe questions developed from a list of 20 subtopics on snakes; (3) a word association task based on the list of the 20 subtopics about snakes; (4) a recognition task, based on a modified multiple choice format, about snakes; and (5) a relatively unstructured discussion of the child's first experience with snakes. Each subject met individually with the experimenter in five separate sessions. The structured questions elicited the greatest quantity of information or number of facts, followed by the recognition task, the word association task, free recall task, and the unstructured interview. The structured question condition produced both more information and more correct information than any other task. (MKM)