Amazonia is a megadiverse tropical forest biome of continental dimensions. Although still largely intact, the biome is highly threatened and future infrastructure projects associated with the low governance in Amazonian countries suggest a bleak scenario for the region in years to come. Biodiversity conservation in Amazonia has focused on creating a robust protected area (PA) network to face human-related threats. At present, some 34 % of the hydrologic domain of the entire Amazon Basin is under some category of protection. However, total acreage may not be the best way to assess levels of protection against anthropogenic pressures. This thesis considers the vulnerability to anthropogenic threats and the distribution of PAs throughout Amazonia at the scale of major watersheds. Watersheds are presented not just as a scale of analysis but also as a viable option for conservation planning units across Amazonia. Analysis of 23 major watersheds indicated that high-vulnerability are widespread all over Amazonia. However, the most threatened areas are located in the southeastern and western portions of the biome and there is clear evidence of a mismatch between PAs and high vulnerability areas. This imbalance in the allocation of conservation investments within the biome leads to either under-protection or over-protection, creating redundancy in setting aside similar habitats, thereby misusing scarce available resources. In addition, the systematic PA avoidance of high vulnerability areas is a strategy that is both deceptive and risky, because the future prospects of biodiversity conservation performance of existing hinterland reserves are less than sanguine once they eventually confront severe threats from advancing development frontiers. Creating PAs far from high pressure areas is a reasonable strategy to meet global conservation goals, but often merely serve to justify political objectives with questionable impacts on biodiversity protection.