Background and aims: Evidence suggests that gait in complex situation is influenced by executive function in normal ageing. But to our knowledge, no study revealed what measurement is the most appropriate to assess impaired walking in older adults. The objective is to (1) test the feasibility of our protocol and (2) to explore the cognitive load during a simple vs. dual walking task in younger and older adults. Methods: To date, 6 young adults (mean = 26.7 years old; SD = 4.1) were included in this study. Fifteen young adults and 25 older adults will be included by May 2018. Participants are asked to achieve a walking task including four conditions: (1) normal pace walking, (2) subtracting while standing, (3) walking while subtracting and (4) walking while counting forward. Cognitive load is measured by changes in oxygenated and deoxygenated hemoglobin in dorsolateral prefrontal cortex (fNIRS device) and the NASA-TLX questionnaire completed after each condition. Results: fNIRS analyses are in progress. Preliminary results showed that younger adults felt higher cognitive effort in both conditions 2 and 3 compared to condition 1 (NASAscorescondition2 (mean ± SD): 66 ± 18.7 vs NASAscorecondtion1: 9.5 ± 6.3, W = -2.2, P = 0.03 ; NASAscorecondition3 : 78 ± 14.6 vs NASAscorecondition1, W = -2.2, P = 0.03). No significant differences in speed, cadence and stride length were found between conditions 1, 3 and 4 (p > 0.05). No significant differences in subtraction performances were observed between conditions 2 and 3 (p > 0.05). Conclusions: Higher cognitive load in condition 3 in younger adults confirm that our dual task walking is the most cognitively demanding condition. We expect that our fNIRS results in younger adults will be similar to those observed with NASA-TLX questionnaire. However, discrepancies between those two measures are expected in older adults.