Background and Purpose— The goal of this study was to develop a functional MRI (fMRI) paradigm robust and reproducible enough in healthy subjects to be adapted for a follow-up study aiming at evaluating the anatomical substratum of recovery in poststroke aphasia. Methods— Ten right-handed subjects were studied longitudinally using fMRI (7 of them being scanned twice) and compared with a patient with conduction aphasia during the first year of stroke recovery. Results— Controls exhibited reproducible activation patterns between subjects and between sessions during language tasks. In contrast, the patient exhibited dynamic changes in brain activation pattern, particularly in the phonological task, during the 2 fMRI sessions. At 1 month after stroke, language homotopic right areas were recruited, whereas large perilesional left involvement occurred later (12 months). Conclusions— We first demonstrate intersubject robustness and intrasubject reproducibility of our paradigm in 10 healthy subjects and thus its ...