"The purpose of this research is to study forest fire problematic. Economic, statistical and mathematical models are showed in order to manage efficiently prevention and supression forces against forest fires phenomenon. In chapter 1, the phases that characterize forest fire management are explaned. The bibliography about economic valuation of wildfire programs is displayed. Moreover, different evaluation methods of losses caused by forest fires in no market forest resources are discussed. In chapter 2, forest area, number of forest fires, hectares burned, causes and economic effects of wildfires are described in the context of European Union and Spain. The general increase of forest fires occurence in Spain has implied the need of better management programs against wildfires in order to solve the problem. The development of models that forecast forest fire danger (chapter 3) is considered as one of the most important elements in order to dispatch the neccesary firefighting resources. Therefore, chapter 3 deals with models to estimate and to predict people-caused forest fires. Some of the models proposed are new in forest fire occurrence literature, such as Gompertz Regression, Extreme Value Regression, Flexible Logistic Regression and ARIMA models. Chapter 4 presents the estimations and predictions, coming from the statistical models, of historical daily forest fire occurence for a concrete area. Finally, in chapter 5, an model of allocation of resources is proposed. For this model, it is neccesary to take into account forest fire occurence estimation, price of forest outputs and range of hectares burned. In this context, the model can estimate the quantity of neccesary forest resources in order to satisfy the level of outputs stablished by manager for a especific area."