Even if optical correlates of self-motion velocity have already been identified, their contribution to the control of displacement velocity remains to be established. In this study, we used a virtual reality set-up coupled to a treadmill to test the role of both Global Optic Flow Rate (GOFR) and Edge Rate (ER) in the regulation of walking velocity. Participants were required to walk at a constant velocity, corresponding to their preferred walking velocity, while eye height and texture density were manipulated. This manipulation perturbed the natural relationship between the actual walking velocity and its optical specification by GOFR and ER, respectively. Results revealed that both these sources of information are indeed used by participants to control walking speed, as demonstrated by a slowing down of actual walking velocity when the optical specification of velocity by either GOFR or ER gives rise to an overestimation of actual velocity, and vice versa. Gait analyses showed that these walking velocity adjustments result from simultaneous adaptations in both step length and step duration. The role of visual information in the control of self-motion velocity is discussed in relation with other factors.