Despite motion of the entire retinal image that results from fixational eye-movements, the visual scene is perceived as stationary. One hypothesis to account for this observation is that normal motion sensitivity is limited by the variability of fixational eye velocity. The present experiments tested this hypothesis by comparing motion sensitivity and the variability of fixational eye velocity in corresponding meridians. Speed thresholds to detect horizontal, vertical, and rotary motion in a set of eight random-dot patches were measured, while normal observers monocularly viewed the stimulus with gaze either straight-ahead or deviated to the left by 45°. Eye-movement recordings using the search-coil technique were used to estimate the variability of eye velocity in the horizontal, vertical, and torsional meridians during fixation. As reported previously by Murakami (2004), the averaged thresholds for horizontal and vertical motion correlated with the averaged variability of eye velocity in the horizontal and vertical meridians when observers looked straight-ahead. However, no relationship existed between the threshold for rotary motion and the variability of eye velocity in the torsional meridian. Furthermore, no relationship existed between the motion threshold and the variability of eye velocity in any meridian during fixation in lateral eccentric gaze. These results are only partly consistent with the hypothesis that fixation variability limits motion sensitivity.