A study examined the relationship between types of miscues made by readers during oral reading and the errors made by the same readers on a cloze task. Subjects were 79 third and 78 fifth grade students in eight intact classrooms. Both the oral reading miscues and the cloze replacements were categorized using a modification of Y. Goodman's miscue analysis system. The cloze tests were prepared by leaving the introductory sentence of a passage complete and thereafter deleting every fifth word. The reading passages were grade appropriate and selected from naturally occurring instructional materials. Each subject was tested in two sessions, with both the cloze test and a standardized reading test administered in the classroom setting. The oral reading sample was collected during an individual session with each subject. Results indicated that there were clearly some strong similarities between information yielded by a miscue analysis and a relatively indepth exploration of the types of responses made on a cloze task. In general terms, oral reading accuracy was closely related to success in completing the cloze tasks. The two classification systems, however, frequently produced disparate reading level estimates for individuals within the sample. (FL)