We actively scan the environment around us by moving our eyes. Unlike cameras, our eyes are never held still, even during fixation. These fixational eye movements (FEMs) are not the result of imperfect control of our eyes, but instead, they are fine and exquisite movements controlled by our oculomotor system. While we now know that FEMs serve as essential computational steps for visual processing of fine details, it remains unclear how they are planned or exert influence on our visual perception. Here, we investigated the role of FEMs in active visual processing and perception using a multidisciplinary approach incorporating computational modeling, human psychophysics, and non-human primate physiology. To study the role of FEMs in visual perception, I built a computational model and discovered that certain FEM trajectories are more advantageous than the others during visual discrimination. The modeling results made predictions about the neurophysiological consequences of FEMs in early visual pathway. In psychophysical experiments making use of high-precision eye tracking, I found that ocular drifts – the small, persistent, and seemingly involuntary eye movements made during fixation – are under cognitive influence. In neurophysiological studies in non-human primates, I found that V1 neural activity is tuned by the visual information acquired by microsaccades and ocular drifts, and that the modulation depends on the spatial frequency of the visual stimuli. In summary, I identified trial-by-trial influence of FEMs, task knowledge influence on FEMs in visual discrimination, and modulation of V1 activity by reafference during microsaccade-drift cycle. Together, these results help to delineate the role of the visuomotor control of FEMs in active sensing.