To determine the most accurate group-administered measure of reading achievement, a study explored variations of the cloze and maze procedures with second grade students who were native English speakers or who were being taught English as a second language. Subjects--108 second grade volunteers (1% American Indian, 49% Asian, 39.8% Black, 1% Hispanic, and 9.2% White) in six classrooms in an urban elementary school--were familiarized with both cloze and maze formats. Two cloze passages, two maze passages, and the reading sections of the Comprehensive Test of Basic Skills (CTBS) were then administered in a counter-balanced sequence within a three-week period. Subjects also indicated the appeal of each passage by making a smiling, neutral, or frowning face. Performances were compared by the language background of each participant. Analysis of results showed that for all subjects the cloze was more accurate than the maze procedure; it more closely reflected the subjects' reading performance than did the CTBS. (A list of references and tables of data are included.) (AEW)