Aims: The aim of this study was to examine physical activity and sedentary behaviour characteristics of ambulatory and community-dwelling patients with dementia compared to cognitively healthy age-, sex- and weight-matched controls. Methods: In this cross-sectional study, we included community-dwelling dementia patients (n = 45, age 79.6 ± 5.9 years, Mini-Mental State Examination [MMSE] 22.8 ± 3.2) and matched controls (n = 49, age 80.0 ± 7.7 years, MMSE 29.0 ± 1.2). Participants wore a wrist accelerometer for 7 days to assess sedentary time, sedentary bout duration and time spent in very light, light-to-moderate and moderate-to-vigorous physical activities. Results: Relative sedentary time and sedentary bout duration was significantly higher in dementia patients than in controls (median [interquartile range] 57% [49–68] vs. 55% [47–59] and 18.3 [16.4–21.1] min vs. 16.6 [15.3–18.4] min, p = 0.042 and p = 0.008, respectively). In addition, dementia patients spent a lower percentage of their waking time in light-to-moderate and moderate-to-vigorous intensity physical activities (20% [15–23] vs. 22% [18–25] and 5% [2–10] vs. 10% [5–13], p = 0.017 and p = 0.001, respectively). Conclusion: We revealed that dementia patients are more sedentary and perform less physical activity than cognitively healthy controls. This may have clinically important consequences, given the observation that sedentary behaviour and little physical activity independently predict all-cause mortality and morbidity.