The human pulvinar has been subject to increasing study as evidence accumulates that it plays an active role in visual perception, specifically in spatially directed attention. Single-unit studies have reported retinotopic organization in some sub-regions of the monkey pulvinar (Bender, 1981). However, such topography has not been found in human subjects. Establishing whether topographic encoding exists is a fundamental step in advancing our understanding of the pulvinar's role in visual processing. Here we use fMRI to test for position-sensitive encoding in the human pulvinar. Subjects passively viewed flickering Gabor stimuli that were located at one of five eccentricities from a central fixation point. Spatial patterns of BOLD responses corresponding to each of the Gabor positions were cross-correlated, producing a Fisher z score for each pair. As the spatial separation between Gabors increased, the correlation between patterns of activity across voxels within the right pulvinar decreased significantly, indicating precise position-sensitive encoding. Right hemisphere pulvinar discriminated Gabors separated by less than 2 degrees visual angle (at ∼10 deg. eccentricity). The voxels within the pulvinar showing the strongest position sensitivity exhibited more precise discrimination than did 99.99% of the voxels in the rest of the brain outside of visual cortex (V1, V2, V3, V4, and VP). The data also reveal a surprising lateralization: across all seven subjects, right hemisphere pulvinar showed strong topographic encoding while left hemisphere pulvinar showed little or no position sensitivity. Other studies have provided some evidence for lateralization of function in the pulvinar, e.g. Karnath et al. (2002), in which a group of stroke patients suffering from spatial neglect had damage to the right pulvinar but not the left. It is possible that the lateralization in topographic encoding that we see here is related to the functional lateralization reported by Karnath and elsewhere.