Recent findings of a high response to bodies in the fusiform face area (FFA) challenge the idea that the FFA is exclusively selective for face stimuli. We examined this claim by conducting an fMRI experiment at high resolution (1.4 x 1.4 x 2.0mm) on 9 subjects using visual stimuli in both blocked and event-related designs. Regions of interest (ROIs) were defined using data from the blocked-design runs, during which subjects viewed images of faces, headless bodies, and objects. We identified the FFA as the face-selective region on the fusiform gyrus with greater activation (p < 0.0001) for faces versus objects and an adjacent body-selective region with greater activation (p < 0.0001) for bodies versus objects (see also Peelen & Downing, 2004). These regions overlapped in all subjects. To test whether separate and exclusive selectivities exist for faces and for bodies, we excluded the dual-selectivity voxels from further analysis, thereby creating two new ROIs, one that showed significant face but not body selectivity in the localizer, and one with the opposite pattern of body without face selectivity. On average 75 percent of voxels in the FFA were exclusively face-selective and 56 percent of the voxels in the body-selective ROI were exclusively body-selective. Our event-related data from the same subjects replicated this exclusive selectivity in each ROI: the faces-only region produced a significantly higher response to faces (1.04 PSC) than to bodies (0.44 PSC) or objects (0.39 PSC), which did not differ from each other. Conversely, the bodies-only region produced a significantly higher response to bodies (0.90 PSC) than to faces (0.55 PSC) or objects (0.56 PSC), which did not differ from each other. These results demonstrate strong and exclusive selectivities in distinct but adjacent regions in the fusiform gyrus for only faces in one region and only bodies in the other.