In artificial laboratory situations where subjects undergo repetitive self-generated or externally imposed head rotations, visual-vestibular interaction during the wearing of telescopic spectacles can markedly augment gain of the vestibulo-ocular reflex (VOR). The present study was conducted to determine whether the wearing of these aids for the visually impaired is associated with similar visual-vestibular interaction during more natural activities. Angular eye and head movements of unrestrained normal volunteers were measured using magnetic search coils. In some subjects, head translations and rotations were also monitored by a flux gate magnetometer array. Measurements were performed of the VOR in darkness, and of the visually enhanced VOR (VVOR) in lit conditions, during three natural activities: 1) standing quietly; 2) walking in place; 3) running in place. These data were compared with similar measurements during repetitive voluntary head oscillations at 0.8 Hz in pitch or yaw. During VVOR, subjects viewed a target placed 6 to 10 m away and remembered this target during VOR trials in darkness. To assess the effects of altering visual-vestibular interactions, VVOR testing during normal vision was augmented by wearing of binocular telescopic spectacles of 2X, 4X, and 6X powers. Dorsoventral and mediolateral head translations were consistently phase-locked with pitch and yaw head rotations, respectively, such that head translation at least partially compensated for rotational disturbances of gaze. Angular velocity of the head was greater during walking than during standing, and was greater still during running, with a greater increase in each case for pitch as compared with yaw. Eye movements were phase compensatory for head movements. VOR gain (eye velocity divided by head velocity) was near 1.0 in both pitch and yaw during standing and during actively generated head rotation. During walking and running there was a significant decrease in angular VOR gain in pitch to approximately 0.75 (P < 0.0005). During ambulatory activities, normal and magnified vision were associated with VVOR gain enhancement in pitch and yaw that was statistically significant, but substantially less than was telescope magnification and markedly lower than was the corresponding VVOR gain measured during active head rotation. Measurements of unmagnified VOR and VVOR during walking and running showed that gain was lower than the "ideal" value of 1.0. However, since translational head perturbations during these activities partially offset the visual effects of angular disturbances, lower gains may nevertheless be associated with retinal image stability at typical indoor target distances. In contrast with performance during repetitive, uniplanar motion, vision has very limited influence on VOR gain during natural activities.