Selective attention models predict that verbatim study questions can divert students from meaningfully encoding attributes of science concepts. The aptitude-treatment interaction (ATI) hypothesis predicts that such questions can be particularly dysfunctional to low-ability students. These predictions assume the measurement of true comprehension of concepts as a criterion. Eighth-grade students (n = 217) were randomly assigned to a text-only, text-question or a placebo treatment. The text verbally described five fossil types. The questions consisted of 28 fill-in-the-blank queries about the text. The posttest required students to visually identify and discriminate 40 fossil specimens as to fossil type. Comprehension of the concepts clearly took place-a fact substantiated by the very low scores obtained by the placebo group. As predicted (p < 0.05), low-verbal students performed better when provided with a text-only rather than a text-question treatment. In contrast, high-verbal students were less effected by the verbatim study questions. Main effects among these groups were also detected. Apparently such questions can overprompt students, resulting in their copying of words from a text to an answer-blank without semantically encoding (i.e., comprehending) the copied words.