Background: Greenspace exposure is associated with improved childhood development, yet the pathways behind this relation are insufficiently understood. Here, we investigated the association between lifetime residential exposure to greenspace and early childhood development and evaluated the extent to which this association was mediated by reductions in traffic-related air pollution and noise. Methods: In a large population-based birth cohort in Metro Vancouver, Canada (n=27,593), early childhood development was assessed via teacher-ratings on the Early Development Instrument (EDI) (mean age = 5·6 years). Greenspace was estimated using vegetation percentage derived from spectral unmixing of annual Landsat satellite-image composites. Lifetime residential exposure to greenspace (from birth to time of EDI assessment) was estimated as the mean of annual vegetation percentage values within 250 m of participants’ residential postal code. Multilevel modeling was used to investigate associations between greenspace exposure and EDI scores. The mediation effects of NO2, PM2.5, and noise levels were estimated using causal mediation analyses. Findings: In adjusted models, one interquartile range increase in vegetation percentage was associated with a 0·16 increase in EDI score (95% CI: 0·03, 0·29), indicating small improvements in early childhood development. 86·6%, 20·6%, and 37·3% of the association was mediated through reductions of NO2, PM2.5, and noise exposure levels, respectively. Interpretation: Increased residential greenspace exposure may improve childhood development by reducing the adverse developmental effects of traffic-related exposures, especially NO2 air pollution. This study contributes to knowledge of pathways between greenspace exposure and childhood development and supports healthy urban planning. Funding Information: This work was supported by the Canadian Institute of Health Research (grant reference number: 156152). Declaration of Interests: We declare no competing interests. Ethics Approval Statement: Ethics approval for this study was granted by the University of British Columbia Behavioral Research Ethics Board (certificate H18-00908).