INTRODUCTION: Moderate to vigorous physical activity (MVPA) is linked to lower dementia risk, though the optimal timing remains unclear. METHODS: Using accelerometer-measured MVPA data from 90,329 dementia-free UK Biobank participants with at least 2 years of follow-up, we assessed associations between MVPA patterns (morning, afternoon, evening, mixed, inactive) and dementia risk via multivariable-adjusted Cox proportional hazards models. RESULTS: Compared to the inactive group, participants with a tendency of morning MVPA had a lower risk of all-cause dementia (hazard ratio [HR]: 0.60, 95% confidence interval [CI]: 0.40-0.90) and Alzheimer's disease (AD) risk (HR: 0.44, 95% CI: 0.22-0.86). No association of the other time point with all-cause dementia was observed. Additionally, the incidence of vascular dementia (VD) did not exhibit a significant relationship among all timing groups. DISCUSSION: Morning MVPA was associated with a lower risk of all-cause dementia and AD, while MVPA at any time point showed no association with VD. Highlights Few studies have examined accelerometer-measured moderate to vigorous physical activity (MVPA) timing with incident dementia. Participants with a tendency of morning MVPA had a lower risk of all-cause dementia and Alzheimer's disease. No significant association was found between all MVPA timing and vascular dementia (VD) risk. Findings suggest timing-specific physical activity interventions for dementia prevention.