A method for measuring sulcal and gyral patterns, using data derived from magnetic resonance (MR) scanning, is described. This method can be applied through two newly developed computer programs, BRAINPLOT and BRAINMAP. These programs provide quantitative measures of brain surface pattern. The method has been validated with postmortem brains, phantoms, and human MR data. The method is robust to detecting differences in brain surface anatomy between atrophic and nonatrophic brains. It appears to offer an efficient, fully automated, and accurate method for analyzing the large amounts of information generated through in vivo neuroimaging techniques.