Aphasia is a common outcome following brain injury and often leads to long-term disability and diminished quality of life. Language production disorders in particular, such as difficulty finding the right word or problems speaking in sentences, are the most common symptoms of aphasia, and significantly hinder communication and social participation. While some people with language production deficits recover well during the first year after injury, others will not, and this discrepancy in outcomes is largely unpredictable. Many factors affect recovery patterns across individuals, one of which is the involvement of different neuroanatomical structures in the resulting impairment and subsequent degree of recovery. Surprisingly, little is known about structural changes in the brain accompanying recovery from language production deficits in aphasia, though preliminary evidence suggests they may be better predictors of recovery than traditionally-used functional change measures. The proposed longitudinal, multimodal neuroimaging study will explore the patterns of recovery in a large group of individuals with aphasia following stroke. By combining a number of novel techniques, we will investigate the contributions of numerous gray and white matter structures to the recovery of language production deficits. Fifty individuals with aphasia will be tested at 1-, 3- and 12-months post stroke on various measures of language production at the word, sentence and discourse level to establish baseline measures of performance and monitor changes across the different stages of recovery. At the same three time points, participants will undergo structural MRI and high angular resolution diffusion imaging (HARDI) to determine lesion site, white matter integrity, and changes in grey and white matter over time. The first main objective of the study is to determine which damaged cortical regions and white matter pathways impede recovery of language production throughout the first year. The second main objective is to measure neuronal changes in gray matter morphometry and white matter integrity that accompany observed behavioral improvement. Evaluating different time periods during the first year post-stroke will enable us to tease out neural indices predicting, impeding, and supporting language recovery at different stages. The proposed project will be the first comprehensive longitudinal study using state-of-the art techniques to examine plasticity within gray matter regions and white matter tracts and their contribution to recovery of different language levels over the course of one full year. In the end, the results of this study will provide important insights into factors contributing to successful recovery of language abilities and enable clinicians to utilize neuroimaging information in prognosis and selection of the most optimal treatments.