A free response cloze test was administered to 31 third graders to test the hypothesis that qualitative as well as quantitative differences in syntactic abilities or skills exist between good and poor readers. The subjects were divided into two reading level groups (good readers and poor readers) according to their scores on the SRA Assessment survey after their second grade year. A second cloze passage was administered to the same subjects and to an additional group of 13 very poor readers from another school. The results of the two measures suggest that poor readers make more errors that violate the grammatical structure of the sentences in which those errors occur, indicating that poor readers are less sensitive to, or less able to use, syntactic information than good readers. The study concludes that, if these differences between good and poor readers can be replicated under other conditions, then alternate instructional procedures for poor readers should be explored. (RL)