Recovery of language in stroke-induced aphasia is little understood, with extant theories suggesting that ipsilesional and/or contralesional tissue may be recruited into the language network. However, several cognitive and neural factors contribute to recovery patterns, including language deficit, lesion volume/location, and neurophysiological variables. In this talk I will present first-phase findings derived from the Center for the Neurobiology of Language Recovery, a multisite NIH funded project focused on recovery of anomia (Boston University, Harvard MGH), dysgraphia (Johns Hopkins University) and sentence processing (Northwestern University), respectively. Results of cross-site projects will be presented, including BOLD signal patterns and activation reliability, perfusion patterns, and right-hemisphere density found in over 50 chronic stroke aphasic participants. The neurobiology of recovered sentence processing also will be discussed, emphasizing the effects of treatment versus natural history on sentence processing deficits on behavioral (off-line and on-line (eyetracking) measures) and neural processing (BOLD signal).