The authors tested the hypotheses that neuropsychological functioning would improve after homeless persons with severe and persistent mental illness were provided with housing and that executive functioning would improve more among those placed in group homes than among those placed in independent apartments. A total of 114 persons with serious and persistent mental illness who were stable residents of homeless shelters completed neuropsychological testing and were randomly assigned to group homes or independent apartments; 91 participants (52 assigned to group homes and 39 assigned to independent apartments) were retested after 18 months. (abstract from article)