Image analysis software allows for quantitative estimation of widths, areas, and/or volumes of central nervous system (CNS) structures directly from digital images. Although atrophy is not pathologically specific, it primarily reflects irreversible tissue loss due to multiple sclerosis (MS), and therefore, it is a valuable marker of disease severity. Brain atrophy can be detected very early in the course of MS, and appears to progress almost from disease onset. Current evidence suggests that atrophy correlates better with neurologic measures of disability than do conventional lesion measurements. Atrophy is an attractive component of a magnetic resonance imaging (MRI)-based outcome assessment in MS clinical trials because it reflects diffuse pathologic processes that are not accounted for by lesion measurements, and yet it can still be measured from images acquired with conventional MRI pulse sequences. Gray matter atrophy may provide a feasible measure of the extent of cortical pathology.