BACKGROUND: Dual-task-based assessment tests failed to establish a dependable relationship between dual-task-related gait changes and the risk of falls in the elderly. OBJECTIVE: The aim of this study was to examine whether changes in gait while counting backward could be associated with the occurrence of a first fall among older adults. METHODS: Walking while counting backward was investigated prospectively in a cohort of 187 older adults living independently in senior housing facilities. During enrollment, walking time, number of steps, and frequency of lateral line stepping-over and stops were measured while walking only and while walking with backward counting aloud. Information on the incident falls during the follow-up year was collected monthly. RESULTS: Walking time and the number of steps increased significantly under the dual-task condition compared to the single-task condition among fallers and non-fallers (p < 0.001). Compared to non-fallers, fallers had significantly lower scores in the Mini-Mental State Examination (p = 0.029) and higher scores in the 15-item Geriatric Depression Scale (p = 0.003) and Timed Up & Go Test (p = 0.006) and increased walking time under both walking conditions (p = 0.030 for single-task condition and p = 0.007 for dual-task condition). After adjusting for these variables, depressive symptoms (adjusted OR = 2.6 with p = 0.041 and adjusted OR = 2.5 with p = 0.045 when walking time while walking only and walking with backward counting is considered, respectively) and walking time while walking only (OR = 2.3 with p = 0.032) were significantly associated with falls. CONCLUSION: Dual-task-related gait changes were poorly associated with the occurrence of a first fall and provided no additional predictive value compared to gait performance under a single task, suggesting that changes in basic clinical gait parameters while counting backward are unsuccessful to predict the first fall among older adults.