In urban areas, estimating the effect of land cover (LC) data spatial resolution on ecosystem services (ES) mapping remains a challenge. In particular, mapping spatial flows of ES, from greenspaces to beneficiaries, may be more sensitive to LC data resolution than mapping potential supply or demand separately. Our objectives were to compare the sensitivity of global- and local-flow ES maps to LC data resolution, and to assess the effect of LC data resolution within different types of urban land uses. A case study was conducted in the city of Laval, Canada. Carbon storage (a global-flow ES), urban cooling and pollination (two local-flow ES) were mapped using LC data aggregated from 1 to 15 m. Results were analyzed for districts (comprising various types of urban land uses), and for 480 x 480 m residential and commercial zones. Greenspace cover was generally underestimated at coarser spatial resolutions; as a result, so were ES potential supply and flow. For urban cooling and pollination, the effect of LC data spatial resolution on ES flow also depended on changes in the spatial configuration of ES potential supply relative to ES demand. The magnitude of the effect differed among land use types. However, the effect was also highly variable between similar landscapes, suggesting that it is very sensitive to LC structure. To adequately map the ES provided by the small greenspaces scattered throughout the urban matrix, using land cover data with a spatial resolution of 5 m or finer is recommended, especially for local-flow ES.< Réduire