This study investigates how German which-questions are processed in cochlear implant (CI) simulated speech compared to normal speech. In addition, it explores the role of working memory capacity (WM). Using eye-tracking, accuracy and gaze patterns for subject, object and passive questions were studied. Questions were disambiguated with grammatical information, specifically morphosyntactic cues: case and/or subject-verb agreement. Results show that only object question accuracy was affected by CI simulation. Similarly, eye movements showed weaker interpretation preferences for object questions in CI simulation. Participants with high WM had higher accuracy and their eye gaze showed quicker revisions of garden path effects compared to those with medium or low WM capacity. These findings show that morphosyntactic cues support correct interpretation of object questions in CI simulation, but processing these questions is slower and influenced by working memory capacity.