Humans typically use medio-lateral movements of the fingers to explore surfaces. In such a movement, the finger pad is compressed against the surface, stretched, then slides over the surface, eliciting high frequency vibrations, and is finally unstretched. This cascade of mechanical events stimulates mechanoreceptors in the skin, which transform physical events into a neural code, informing us about the qualities of surfaces/objects. Neurophysiological studies performed to define the neural code at peripheral level have described how the spatial details of textures (e.g. Braille dots) or the vibratory aspects of the skin/surface interactions are coded in animals and humans using stimuli applied passively. We aimed to investigate how the dynamics of a naturally-produced lateral movement over a texture shapes the mechanoreceptive afferent responses to deliver a meaningful touch percept. Using microneurography in humans, we recorded neural responses of mechanoreceptors during active touch of a variety of natural textures, at different speeds of exploration. Our results reveal a specific pattern of neural responses within movements, with the slow adapting mechanoreceptors (SAs) responding optimally at movement initiation/termination, and fast adapting mechanoreceptors (FA) responding optimally during sliding. Furthermore, in a subset of afferents, we correlated firing with physical and perceptual descriptors. These relationships suggest that a subset of SAs (SA2) code the friction coefficient of the texture, while FAs code the vibratory aspects. Finally, FA firing increased with faster movement speed, while SAs were relatively insensitive to changes in speed. These findings expand current knowledge on the neurophysiology of touch in humans.