Forest fires produce a major impact on soil, water and vegetation. Despite the amount of research published on this subject, there are two major problems that hamper the fully understanding of on and off-site impacts of forest fires. They include methodological problems steaming from the uniqueress of burned soil properties, easily erodible, by the fast degradation they undergo over a short period of time immediately after fire and by the meaning of the impacts at different scales.Monitoring attempts to understand processes in burned areas are hindered by limitations of measuring techniques, that prevent the correct quantification of erosion yields and the processes that give rise to peak flows. A further limitation arises from the poor knowledge on how properties and processes at one scale influence degradation processes at larger scales, both on and off site.This paper presents a reflection about the limitations of some of the methods and techniques more frequently used to assess erosion yields and hydrological responses following fires, and their significance at different scales of analysis.It also shows the potential of nested approaches in the acquisition of an improved insight in to the problem and in the identification of the relevant processes at each scale and how they influence degradation processes at larger areas. It is shown that soil and land use patterns, play a crucial role in reducing or enhancing the hydrological and sediment yield and transport processes between scales. © 2008 Elsevier B.V. All rights reserved.