Background: Although walking without vision seems to carry a high cognitive cost, few studies have measured the cognitive load involved in this activity in blind people. The aim of this study was to assess the cognitive load of walking in blind people, using gait analysis, a dual task paradigm and a subjective assessment of cognitive load. Methods: In a quantitative quasi-experimental design, 25 blind adults walked 40 meters. In one trial, participants walked normally (control condition). In another, they walked while performing an auditory simple reaction time task, and in the third trial they walked, performed the simple reaction time task and avoided obstacles. In addition to the simple reaction time task performance, walking speed was recorded, and participants provided a subjective assessment of cognitive load after each trial. Performance of participants aged less than 60 years were compared with those aged over than 60 years. Results: Walking significantly reduced performance of the simple reaction time task; carrying out the simple reaction time task while walking significantly reduced walking performance and increased the subjective assessment of cognitive load; and simple reaction time task performance decreased and subjective assessment increased when obstacles were present. Few significant age effects were found. Significance: Walking without vision involves a cognitive load that increases when the environment becomes complex. Each of the three methods used is relevant when assessing the cognitive load involved in walking in blind people, and could be useful in rehabilitation intervention. The results obtained allowed recommendations to be suggested for the design of technological mobility devices.