PURPOSE Self-reported data suggest that older adults with dementia are inactive. The purpose of the present study was to objectively assess the physical activity (PA) levels of community-dwelling and institutionalized ambulatory patients with dementia, and to compare with the PA levels of cognitive healthy older adults. METHODS We used actigraphy to assess the PA levels in institutionalized (n = 83, age: 83.0 ± 7.6, Mini-Mental-State Examination (MMSE): 15.5 ± 6.5) and community-dwelling dementia patients (n = 37, age: 77.3 ± 5.6, MMSE-score: 20.8 ± 4.8), and healthy older adults (n = 26, age: 79.5 ± 5.6, MMSE-score: 28.2 ± 1.6). We characterized PA levels based on the raw data and classified <100 counts/min as sedentary behavior. RESULTS Institutionalized dementia patients had the lowest daily PA levels (1.69 ± 1.33 counts/day), spent 72.1% of the day sedentary, and were most active between 8:00 and 9:00 am. Institutionalized vs. community-dwelling dementia patients had 23.5% lower daily PA levels (difference M = 0.52, p = .004) and spent 9.3% longer in sedentariness (difference M = 1.47, p = .032). Community-dwelling dementia patients spent 66.0% of the day sedentary and were most active between 9:00 to 10:00 am with a second peak between 14:00 to 15:00. Community-dwelling dementia patients vs healthy older adults' daily PA levels and sedentary time were 21.6% lower and 8.9% longer, respectively (difference M = 0.61, p = .007; difference M = 1.29, p = .078). CONCLUSIONS Institutionalized and community-dwelling dementia patients are sedentary for most of the day and the little PA they perform is of lower intensity compared to their healthy peers. Their highest PA peak is when they get out of bed in the morning. In addition, it seems that institutionalized living is associated with lower PA levels in dementia patients. These are the first results that objectively characterize institutionalized as well as community-dwelling dementia patients' PA levels and confirm that dementia patients are inactive.