It has been suggested that a huge lake, Lago Amazonas, covered a large part of the Amazon basin until as recently as two thousand years ago. According to this hypothesis, the topmost sediments in western Amazonia are almost universally young deposite of lacustrine and deltaic origin. The hypothesis has gained some attention among biologists because of its implications for biological phenomena in Amazonia, especially biogeography and biodiversity. According to the available geological data, however, Amazonia is geologically far more complex than assumed by the lake hypothesis. In the following discussion we will point out the weaknesses of the Lago Amazonas hypothesis, and indicate alternative explanations of the surface geology that are based on tectonically controlled fluvial deposition.