The effect that student ability level has on receiving feedback following classroom tests was studied. Forty-four undergraduates enrolled in four educational psychology classes were assigned to low or high ability groups based on their total score from the first four exams. Two classes were trained in a feedback technique, and the remaining two classes served as the control. One class meeting following each exam was used as a feedback session. All students were given their scored answer sheets and a copy of the exam and asked to review their exam. Students were told to first review items they answered incorrectly and to search the text and their notes for the correct answer. They were then to review items they answered correctly and to review the text concerning items about which they were uncertain. Students in the control sessions were told to review their exams until they were satisfied. All students were administered the same multiple-choice semester tests and the final, which consisted of 40 repeated items, 20 verbatim and 20 paraphrased items, and 10 new items. Only data concerning the 40 repeated items were analyzed. Attention was directed to: the number of correct responses; types of errors for the verbatim and paraphrased items; and new, perseverative, and different error patterns. Findings are discussed. (SW)