Introduction: No evidence exists about the impact of air pollution reduction on incidence of dementia. The aim of this study was to quantify how air quality improvement leads to dementia-incidence benefits. Methods: In the French Three-City cohort (12 years of follow-up), we used parametric g-computation to quantify the expected number of prevented dementia cases under different hypothetical interventions with particulate matter measuring <2.5 μm (PM2.5 ) reductions. Results: Among 7051 participants, 789 participants developed dementia. The median PM2.5 reduction between 1990 and 2000 was 12.2 (μg/m3 ). Such a reduction reduced the risk of all-cause dementia (hazard ratio [HR], 0.85; 95% confidence interval [CI], 0.76 to 0.95). If all study participants were enjoying a hypothetical reduction of more than 13.10 μg/m3 (median reduction observed in the city of Montpellier), the rate difference was -0.37 (95% CI, -0.57 to -0.17) and the rate ratio was 0.67 (95% CI, 0.50 to 0.84). Discussion: These findings highlight the possible substantial benefits of reducing air pollution in the prevention of dementia.