We studied the capacities of human subjects to localize tactile stimuli in 3D space. For that purpose, five subjects (3 women and 2 men) were asked to make a pointing movement to a visual stimulus in otherwise complete darkness. At some random time before, during or after this initial movement, a tactile probe was presented to the tip of the subject's index finger. The probe (1-mm diameter, 0.75-mm extent, 5-ms duration) was applied by a lightweight (80 g) tactile stimulator worn on the subject's hand. To complete the task, the subject was required to point to the 3D spatial location at which the probe was applied. Hand position was monitored (200 Hz) by a video-based motion analysis system. In each subject, probes presented just before or during the initial movement were systematically mislocalized in the direction of that movement so that subjects perceived the probe to be at the location occupied by the hand 50 to 100 ms after probe onset. This mislocalization is likely to be important in dealing with objects in the dark, especially with obstacles encountered along a movement's path. Eye-hand coordination, Spatial localization, Tactile stimulation.