Air pollution is one of the main environmental problems in megacities, such as the metropolitan area of São Paulo (MASP), in Brazil. Urban forests can alleviate air pollution by providing a surface for the dry deposition of particles and trace gases. To benefit from this environmental service and design future green structures, it is crucial to estimate the removal rate of air pollutants by trees. The removal rates of fine particulate matter (PM2.5) by urban trees were quantitatively assessed for the first time in Brazil. A big-leaf modeling approach was adopted, using the i-Tree Eco software. In situ dendrometric data, hourly PM2.5 concentrations, and meteorological variables were used as inputs to the model. PM2.5 removal fluxes ranged between 0.06 and 0.21 g/m2/year in three study areas with contrasting urban forestry conditions. The neighborhood with the greatest canopy cover and tree diversity had the highest removal rates. The evergreen character of the urban forest in the MASP possibly contributed to the relatively high PM2.5 removal fluxes, as compared to other cities around the world. Removal rates were higher in the austral summer, when high precipitation rates restricted the resuspension of deposited particles back to the atmosphere. When extrapolated to the whole metropolitan area, assuming homogeneous forestry conditions, the estimated PM2.5 removal rates were comparable to the magnitude of vehicular emissions, showing that air pollution removal by trees can be substantial in the MASP. The results demonstrate the contribution of urban trees to the improvement of air quality and can boost the development of public policies on urban afforestation in the MASP.