IMPORTANCE: During rehabilitation, stroke survivor's balance is commonly monitored using the Berg Balance Scale (BBS) and Mini-BESTest. These tests evaluate the ability to maintain balance in an upright position while performing tasks. Currently, clinical practice lacks an adequate assessment of balance impairments like the increased postural sway post-stroke. Integrating postural sway measurement with a body-worn Inertial Measurement Unit (IMU) could offer clinicians an objective, and rapid measurement method. OBJECTIVE: Sensitivity to change and responsiveness of postural sway measurement during stroke rehabilitation were explored. DESIGN: A longitudinal study was conducted. SETTING: Data were collected in five rehabilitation centers. PARTICIPANTS: Ninety-four stroke survivors were included in the study. INTERVENTION: Sensitivity to change was evaluated by comparing the percentages of participants who exceeded the Minimal Detectable Change (MDC) for both the conventional and IMU tests. Responsiveness was assessed through hypotheses testing, utilizing both a criterion approach (external criterion: a 5-point retrospective Global Rating of Change score) and a construct approach. RESULTS: At discharge, only 3.2% to 23.9% of the participants showed genuine improvement (ie, had a lower postural sway) on the IMU test, compared to 33% to 60% on the conventional tests. In total, 67.4% of the hypotheses were rejected. CONCLUSIONS: The sensitivity to change and responsiveness of postural sway measurement in stroke survivors are limited. Restoring balance to achieve functional independence is not necessarily dependent on, or associated with, the level of postural sway. At present, integrating postural sway measurement into clinical stroke rehabilitation seems to provide no added value for monitoring balance. RELEVANCE: A novel approach to balance assessment in clinical stroke rehabilitation was investigated. It was found that while most participants improved in terms of functionality and independence, only a small percentage exhibited a significant change in postural sway. Therefore, postural sway may not effectively serve as an indicator of improved movement quality. This insight is crucial for both researchers and clinicians involved in neurorehabilitation.