Background/Objectives: Accurate assessment of physical activity is essential in hospital settings, but difficult for geriatric rehabilitation patients who have slow walking speed and rely on walking aids. This study conducted an external validation of the MOX activity monitor accuracy in distinguishing between sedentary (lying, sitting), standing, and dynamic activities among hospitalized geriatric rehabilitation patients. Additionally, the study assessed clinical tolerance and ease of use of the monitor. Methods: Inclusion criteria were being hospitalized on the geriatric rehabilitation ward, aged 70 years or older, and using a walking aid. The monitor was placed on the upper leg, and remained in place for 1 week. In a test-session, four activities of 4 min each (lying, sitting, standing, and walking) were filmed. Video analyses created a direct observation timeline. Daily skin irritation assessments were conducted and participants noted their discomfort level at the week’s end. Results: The study included 20 participants who relied on walking aids, with 10 requiring additional physical support while walking. The MOX activity monitor exhibited excellent accuracy in classifying the four activities, with a zero mean percentage error for sedentary activity, 2.2% for standing, and 2.7% for walking. Participants reported minimal hindrance, and the monitor proved user-friendly. Conclusion: The MOX activity monitor demonstrated remarkable accuracy in distinguishing between these four activities in a test-session and displayed strong clinical applicability. Significance: The monitor shows substantial promise as a valuable tool for assessing physical activity in hospitalized geriatric rehabilitation patients using walking aids.