This paper presents a computational methodology for predicting the evolution of fire fronts on wild forest landscapes, based on Cellular Automata (CA). The CA framework takes into account various microscopic/detailed and macroscopic factors that affect the spread of fire including the vegetation type, density and height, the wind field, the terrain slope, the spotting fire transfer mechanism as well as suppression tactics. The model was successfully applied to a large scale wildfire that occurred in the mountain of Parnitha, Greece in 2007.