The 3-D multimodality brain image registration algorithm proposed in this paper is based on the matching of feature curves which are defined by the intersections of the interhemispherical fissure plane (IFP) and the skull surface. The IFP is detected by a principal axes technique while the skull boundary in each scan is extracted by tracing the outermost closed contour. The feature curve are rotated to lie on 2-D planes which are parallel to each other so that the problem is reduced to a 2-d curve matching task. MRI-PET image registration result is presented.