Purpose: This study investigated whether sentence repetition–based working-memory (SR-WM) treatment increased sentence-repetition abilities and the treatment effects generalized to sentence-comprehension abilities, WM-span tasks, and general language-assessment tasks. Method: Six individuals with aphasia participated in the study. The treatment consisted of 12 sessions of approximately 1 hr per day, 3 times per week. The SR-WM treatment protocol followed components including maintenance and computation of linguistic units by facilitating a chunking strategy. We manipulated the length and syntactic structures of the sentence-repetition stimuli using a limited set of vocabulary. Results: Participants demonstrated significant increased repetition ability in treated and untreated sentences after treatment. Furthermore, they showed generalization effects on the sentence-comprehension task, WM measures, and general language tasks, but with some differential patterns, depending on task demands. Conclusions: The SR-WM treatment approach, by manipulating syntactic structures and minimizing top-down semantic processing, elicited increased performance on sentence repetition as well as other linguistic domains. Results indicated that it is clinically and theoretically important to examine whether WM treatment serves as a potentially underlying treatment approach that facilitates the distributed network associated with language processing.