This study compared the ability of language learning disabled children and sex/age matched normals to judge the correctness of linguistic stress. Subjects were presented with prerecorded pairs of question-answer trials. In one series they were asked to judge the appropriateness of linguistic stress for each pair. In a second series, they judged semantic appropriateness of the pairs in order to provide a linguistic point of reference for their understanding of the questions. An analysis of variance indicated that the linguistic stress task was more difficult than the semantic interpretation task (p less than .001) and that normal children performed significantly better than LD children on both tasks (p less than .05). Discussion of the data includes interpretation from both perceptual deficit and symbolic deficit perspectives.