Kristi A. Clark, Ph.D., University of California, Los Angeles, hopes to further understanding of the connectivity patterns in autism. One of the current dominant hypotheses is that autism can be characterized by an abnormal neuroanatomical connectivity pattern. Specifically, the hypothesis is one of local overconnectivity combined with long-distance underconnectivity. Graph theory methods, a form of mathematical modeling, have recently been applied to study of structural neuroanatomical connectivity using high angular resolution diffusion imaging (HARDI) data, which allows for the quantification of short-distance and long-distance connectivity. Dr. Clark and colleagues will test these methods in a pilot trial with autistic and nonautistic eight-year-old boys.