Abstract. The present study investigates the radiative effects of dust aerosols in the Mediterranean region during summer 2012 using a coupled regional aerosol–atmosphere–ocean model (CNRM-RCSM5). A prognostic aerosol scheme, including desert dust, sea salt, organic, black-carbon and sulphate particles, has been integrated to CNRM-RCSM5 in addition to the atmosphere, land surface and ocean components. An evaluation of this aerosol scheme of CNRM-RCSM5, and especially of the dust aerosols, has been performed against in situ and satellite measurements, showing its ability to reproduce the spatial and temporal variability of aerosol optical depth (AOD) over the Mediterranean region in summer 2012. The dust vertical and size distributions have also been evaluated against observations from the TRAQA/ChArMEx campaign. Three simulations have been carried out for summer 2012 with CNRM-RCSM5, including the full prognostic aerosol scheme, only monthly-averaged AOD means from the aerosol scheme or no aerosols at all, in order to focus on the radiative effects of dust particles and the role of the prognostic scheme. Surface short-wave aerosol radiative forcing variability is found to be more than twice as high over regions affected by dust aerosols, when using a prognostic aerosol scheme instead of monthly AOD means. In this case downward surface solar radiation is also found to be better reproduced according to a comparison with several stations across the Mediterranean. A composite study over 14 stations across the Mediterranean, designed to identify days with high dust AOD, also reveals the improvement of the representation of surface temperature brought by the use of the prognostic aerosol scheme. Indeed the surface receives less radiation during dusty days, but only the simulation using the prognostic aerosol scheme is found to reproduce the observed intensity of the dimming and warming on dusty days. Moreover, the radiation and temperature averages over summer 2012 are also modified by the use of prognostic aerosols, mainly because of the differences brought in short-wave aerosol radiative forcing variability. Therefore this first comparison over summer 2012 highlights the importance of the choice of the representation of aerosols in climate models.