Several neurodegenerative diseases, such as Alzheimer's disease, cause atrophy of the cerebral cortex. Measurements of cerebral cortical thickness and volume are used in the quantification and localization of atrophy. It is possible to measure the thickness of the cerebral cortex manually from magnetic resonance imaging, but partial volume effects, orthogonality problems, large amounts of manual labor and operator bias makes it difficult to conduct measurements on large patient populations. Automatic quantification and localization of atrophy is a highly desirable goal, as it facilitates the study of early anatomical changes and track disease progression on large populations. The first step in achieving this goal is to develop robust and accurate methods for measuring cortical thickness and volume automatically. We have developed a new method, capable of both extracting surface representations of the cortical boundaries from magnetic resonance imaging and measuring the cortical thickness. Experiments show that the developed method is robust and performs well on datasets of both healthy subjects and subjects suffering from Alzheimer's disease.