Introduction: Falling during walking is a common problem among the older population. Hence, the challenge facing clinicians is identifying who is at risk of falling during walking, for providing an effective intervention to reduce that risk. We aimed to assess whether the clinical version of the narrow path walking test (NPWT) could identify older adults who are reported falls., Materials and Methods: A total of 160 older adults were recruited and asked to recall fall events during the past year. Subjects were instructed to walk in the laboratory at a comfortable pace within a 6 meter long narrow path, 3 trials under single task (ST) and 3 trials dual task (DT) conditions without stepping outside the path (i.e., step errors). The average trial time, number of steps, trial velocity, number of step errors, and number of cognitive task errors were calculated for ST and DT. Fear of falling, performance oriented mobility assessment (POMA) and mini-metal state examination (MMSE) were measured as well., Results: Sixty-one subjects reported that they had fallen during the past year and 99 did not. Fallers performed more steps, and were slower than non-fallers. There were no significant differences, however, in the number of steps errors, the cognitive task errors in ST and DT in POMA and MMSE., Conclusion: Our data demonstrates slower gait speed and more steps during the NPWT in ST and DT in fallers. There is no added value of DT over the ST for identification of faller's older adults., (Copyright © 2015 Elsevier Ireland Ltd. All rights reserved.)