Human activities have modified a significant part of the tropical forest landscapes across the globe, affecting their ecological characteristics and their capacity to provide ecosystem services. In order to counter act the current decrease in tropical forest quality, avoiding and reversing forest degradation has been included as one of the goals of multiple international agreements; it is of particular importance for the climate change mitigation scheme for Reduced Emissions from Deforestation and forest Degradation, forest conservation and enhancement of carbon stocks known as REDD+. In this thesis, I investigated the challenges and feasibility of measuring and monitoring tropical forest degradation in human modified landscapes, focusing on two types of human activities: shifting cultivation and logging. In Chapter 2, I built a conceptual framework that analyses the applicability of the international definitions of forest degradation, and their contrast with the complexity of tropical forest ecosystems and monitoring capacity in tropical countries. I proposed that given the current data and technological limitations, a quick start option to measure forest degradation is to use a benchmark that can be directly linked with the type and intensity of disturbance processes found in an area. Then in Chapter 3, I further studied disturbance processes by analysing the dynamics of shifting cultivation systems and the use of forest resources by communities. I found through a detailed mapping of high resolution data (10X10 m), that similar amounts of forest cover in tropical dry forests (TDF) were lost and gained between the study period (2004-2010), both at the regional and at the community level. This provides evidence that at least in terms of the above ground biomass pool, shifting cultivation systems in TDF could be considered carbon neutral, which implies that these systems have potential to participate in REDD+. The probability of changes in TDF cover in shifting cultivation systems was found to be dependent on the elevation, slope, amount of TDF available per person within a community, and to the amount of livestock and fence posts used by the communities. The use of forest resources and its relation with forest degradation is further studied in Chapter 4. In this Chapter, I evaluated a series of disturbance indicators that best explain the response of forest attributes to human disturbance, and used these indicators to establish four levels of forest degradation. The feasibility of separating four levels of degradation based on two types of high spatial resolution remote sensing data (SPOT 5 and RapidEye satellite data) was assessed. I found that at the landscape level, based on the use of forest resources it was possible to classify TDF into low and high degradation levels. The capacity to classify the landscape into disturbance levels is further explored in Chapter 5, by using historical logging concessions and multi-temporal time series of medium spatial resolution (Landsat) in tropical moist forest. Through the integration of previous land use information with an analysis of the relation of the amount of green vegetation with respect to soil and shadow that compose a pixel, I determined that almost one third of the forest in the study area has experienced disturbance processes. This work supports the need to advance monitoring of forest cover analysis by incorporating forest condition, which has particular implications for the determination of forest carbon stocks. Overall, this research strengthens the concept that the definition, measuring and monitoring of forest degradation should be tailored to the particular dynamics of disturbance processes; and moreover that a direct link between monitoring capacity of a country and policy formulation is clearly needed to improve tropical forest stewardship.