<p>The world is keeping a close eye on Amazonia, due to its importance as a centre of biodiversity and its essential role in regulating global climate. To ensure the provision of climatic services, landscape and species diversity need protection. However, little is known about Amazonian megadiversity patterns. This lack of information is more acute for squamate reptiles (lizards, snakes, and amphisbaenians). Here we propose a regionalisation and endemism scheme for Amazonian squamate reptiles, based on a taxonomically verified checklist. We recognise 14 areas of endemism and 22 bioregions, determined by range restricted species and exhibiting high spatial heterogeneity. This contrasts species richness patterns, which are apparently defined by spatial homogeneity, without a marked gradient. While there is little support for the river barrier hypothesis in defining the areas of endemism, rivers, together with vegetation, geology, relief, and climatic variables, collectively explain spatial patterns of species composition. Our findings suggest that conservation efforts should include the borders of Amazonia, and create a connected network of conservation areas to protect endemism hotspots harbouring rare and threatened species and to function as a buffer against the expansion of habitat destruction to deeper parts of Amazonia.</p>