The classic cytoarchitectonic maps of the human cerebral cortex (e.g., that of Brodmann [1909]) considerably influenced the concept of localization of function. Although they suffer from drawbacks such as the highly observer-dependent definition of areal borders and the lack of information on intersubject variability of location and size of a cortical area in a spatial reference system, these maps often serve as anatomic references in functional imaging studies. Recent methodologic progress in computerized image analysis of histologic specimens and the development of warping techniques to compensate for intersubject variability of brain structure in the three-dimensional (3-D) space made it possible to overcome these drawbacks. The authors propose a new concept of cytoarchitectonic mapping, which is based on (1) a definition of areal borders using multivariate statistical analysis, (2) a quantitative analysis of similarity and dissimilarity in architecture between cortical areas, and (3) a probabilistic mapping of cytoarchitectonic areas in a 3-D reference system. The combined analysis of architectonic maps and functional imaging data in an identical reference space offers exciting perspectives in the analysis of structural-functional correlations in the living human brain and provides new insights into the architecture of the cerebral cortex.