Twelve developmental dysphasics and twelve control children were tested on De Renzi and Vignolo's Token Test. The performance of the dysphasics, though significantly poorer than that of the controls on Parts 2-5 of the test, seemed to follow a similar pattern. Both groups were errorless on the initial parts of the test. As the demand on auditory retention increased on subsequent parts of the test, the performance of both groups progressively deteriorated. However, the addition of grammatical complexity on Part 5, rather than causing even further deterioration in performance, actually served to improve the performance of both groups, though not significantly. Although the pattern of errors made by the two groups was similar, the quantity of errors made by the dysphasics, as compared to the normal control children, demonstrated the magnitude of the dysphasic children's receptive language impairment. Analysis of errors indicated that dysphasic children have greater difficulty with auditory retention and verbal memory than they do with the grammatical or structural complexity of language. It is suggested that the observed gross language impairment of developmentally dysphasic children does not result, at least primarily, from a specific inability to analyze the linguistic components of language. Rather, the language impairment of these children appears to reflect their primary inability to analyze the rapid stream of acoustic information which characterizes speech and is essential to normal speech perception and language development.