Climate change is affecting the meteorological and climatic conditions in which forest vegetation develops. These conditions are also determinant in present and future forest fire activity. In order to quantify the expected alterations in future fire activity, we built linear regression models using monthly meteorological variables and the recorded fire activity in peninsular Spain. These models were developed for 15 territories and for the number of fires, the forest area burned and the wooded forest area burned. The meteorological variables used as predictors are derived from a Regional Climate Model covering the period 1961–90 as the control scenario and 2071–2100 as the future scenario. In the 13 territories with valid models, the explained variance ranged from 20 to 68% for the number of fires, from 12 to 66% for the forest area burned, and from 12 to 62% for the wooded forest area burned. Based on these models, we calculated the ratios between the estimations for the climatic scenarios SRES A2 and B2 and the estimates for the control period. For the entire area and for the A2/control and B2/control respectively, the ratios are 2.5 and 2 times for the number of fires, 4.6 and 3.4 for the forest area burned and 3.9 and 3 times for the wooded forest area burned. In spite of the uncertainties regarding future climatic scenarios as well as the simplicity of the model, the results nevertheless point to a very significant increase in fire activity. Forest management should therefore be focussed on adaptation procedures as well as on ecosystem resilience. This strategy will promote a more sustainable coexistence of forests and fires in altered fire regimes.