The development of Early Warning Systems in recent years has assumed an increasingly important role in landslide risk mitigation. In this context, the main topic is the relationship between rainfall and the incidence of landslides. In this paper, we focus our attention on the analysis of mathematical models capable of simulating triggering conditions. These fall into two broad categories: hydrological models and complete models. Generally, hydrological models comprise simple empirical relationships linking antecedent precipitation to the time that the landslide occurs; the latter consist of more complex expressions that take several components into account, including specific site conditions, mechanical, hydraulic and physical soil properties, local seepage conditions, and the contribution of these to soil strength. In a review of the most important models proposed in the technical and international literature, we have outlined their most meaningful and salient aspects. In particular, the Forecasting of Landslides Induced by Rainfall (FLaIR) and the Saturated Unsaturated Simulation for Hillslope Instability (SUSHI) models, developed by the authors, are discussed. FLaIR is a hydrological model based on the identification of a mobility function dependent on landslide characteristics and antecedent rainfall, correlated to the probability of a slide occurring. SUSHI is a complete model for describing hydraulic phenomena at slope scale, incorporating Darcian saturated flow, with particular emphasis on spatialtemporal changes in subsoil pore pressure. It comprises a hydraulic module for analysing the circulation of water from rainfall infiltration in saturated and nonsaturated layers in nonstationary conditions and a geotechnical slope stability module based on Limit Equilibrium Methods. The paper also includes some examples of these models' applications in the framework of early warning systems in Italy.