Abstract Background The adverse health impact of air pollution exposure on stroke is already well documented but there is a lack of methodological tools to provide actionable evidence on the potential health benefits of realistic interventions. This study demonstrates the use of a parametric g-computation approach to evaluate the impact of hypothetical interventions targeting long-term exposure to air pollution on reducing the stroke prevalence in Belgium using data from three national Health Interview surveys (BHIS 2008-2013-2018). Methods BHIS data (n = 27536) were linked to environmental data of the participant's residential address. A g-computation approach was used to calculate the potential impact fractions of five air pollution reduction interventions to decrease the prevalence of stroke. Regression models were adjusted for socio-economic, environmental and lifestyle factors. In the first and second scenario, the average annual exposure to PM2.5 was lowered to the WHO guideline (5 μg/m3) and reduced by 25%, respectively. In the third and fourth scenario, the average annual exposure to NO2 was lowered to the WHO guideline (10 μg/m3) and reduced by 25%, respectively. In the last scenario, the average annual exposure to BC was reduced by 25%. Results Stroke was significantly associated with all air pollutants. A reduction in the risk of stroke was observed under the five scenarios: -0.87% [SE:0.24], -0.47% [SE: 0.17], -0.32% [SE: 0.19], -0.16% [SE:0.10], -0.13 % [SE: 0.07]. The proportion of prevented stroke cases would be respectively of 65.5%, 35.4%, 23.9%, 12.3%, 10%. Conclusions This study highlights the importance of air pollution on the stroke burden and demonstrates that air pollution reduction interventions could decrease the prevalence of stroke in Belgium. Key messages • Air pollution reduction interventions could significantly decrease the prevalence of stroke in Belgium. • Major benefits were observed for interventions targeting long-term exposure to PM2.5.