Large-scale atmospheric patterns and near-surface winds are expected to be modified in the future due to climate change, altering the availability of wind resources on a regional scale. These possible changes are especially important in isolated power systems, as is the case for a great percentage of the islands. High-resolution climate regionalization is therefore necessary to assess the future projections of wind resource, mainly in the case of orographically complex territories, such as The Canary Islands. In this work, WRF was used to perform a dynamic regionalization in this Archipelago, using the pseudo-global warming technique to compute the initial and boundary conditions from a reanalysis dataset and from the monthly mean changes obtained from the simulations of fourteen global climate models included in the Coupled Model Intercomparison Project Phase 5 (CMIP5). Projections of mean wind, wind energy density and extractable wind power were obtained for two future decades (2045-2054 and 2090-2099) and for two different greenhouse gas scenarios (RCP4.5 and RCP8.5) and the results were compared with those for 1995-2004. Statistically significant changes in wind resource were found in some areas, mainly during summer. Most of these areas correspond to zones where at present wind farms are located. (C) 2016 Elsevier Ltd. All rights reserved.