Sentence imitation effectively discriminates between adults with and without specific language impairment (SLI). Little is known, however, about the factors that result in performance differences. This study evaluated the effects of working memory, processing speed, and argument status on sentence imitation. Working memory was measured by both a storage and processing and a scope of attention measure. Performance differences on arguments versus adjuncts also tested the predictions of the Procedural Deficit Hypothesis (PDH) of specific language impairment. The PDH predicted that adjunct processing would be more difficult for adults with SLI because it would depend on a deficient procedural memory system. Alternatively, adults with SLI may compensate by engaging an intact declarative memory to process adjuncts. Twenty-three adults with SLI and 23 typical language controls participated in three experiments. First, sentence processing times were recorded for argument and adjunct conditions in a self-paced listening task. Second, working memory and processing speed were measured, and used to predict accuracy in a sentence imitation task. The sentence imitation materials varied in length and argument versus adjunct composition. Finally, procedural and declarative memory systems were assessed for all participants. In the online sentence processing task, there was a processing time advantage for noun arguments as compared to verb adjuncts, but no group difference. Correlations of argument and adjunct frequency with processing times were found for the group with SLI, but only argument frequency effects for the typical language group. Group differences as well as argument differences were found in the sentence imitation task. Participants' working memory also contributed to sentence imitation accuracy, but processing speed did not. Argumenthood, working memory and sentence length interacted, with greater argument status effects in short conditions, and greater working memory effects in long conditions. Both the storage and processing and the scope of attention measures predicted sentence imitation. Declarative memory performance was poorer in the group with SLI than in the group with typical language, whereas there was limited support for a procedural memory deficit in the group with SLI. Overall support for the PDH was mixed. There was evidence of compensatory processing of adjuncts as predicted by the PDH, but the basis for the compensation was unclear given the unexpected poor declarative memory performance. The factors of group classification, working memory and argumenthood were confirmed as factors affecting sentence imitation performance, and the complexity of the task was revealed by interactions of these factors with sentence length. The association of a scope of attention measure to language performance suggested that the storage aspect of working memory was as crucial to adult SLI performance as abilities for concurrent storage and processing or rehearsal. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]