This study tested expectations of the palatable forage hypothesis for Atta sexdens (L.). Literature records on Atta and Acromyrmex nest density in the Neotropics were analyzed and A. sexdens demographical aspects were studied in six transects located along the Colombian Amazon River. Literature research results did not show that Atta and Acromyrmex nest densities are consistently higher in disturbed habitats. In transects, 56 active and 182 inactive nests were found in 2008. Higher active nest numbers and densities were determined for 1992, 1999, and 2006. Nests were present in old forest, in agricultural plot and grassland secondary forest, and absent from open spaces around houses, grasslands, agricultural plots, and a particular sector of old forest. Nest densities were higher in secondary forest. Between 1992 and 2008, the size of disturbed areas increased, whereas nest numbers and densities diminished. Average nest age was higher in secondary forest, but maximum nest age was comparable in the three habitats. On average, nests were superficially larger in old forest. Nests were mostly established in old forest, at sites without fallen trees. Almost all of those currently present in secondary forest were established in old forest before any disturbance occurred. In disturbed habitats, nest inactivity was slightly higher and possibly affected by human control efforts. This study suggested that the palatable forage hypothesis alone did not explain observed demographic patterns. Soil physiochemical variables or microclimatic conditions also are thought to play important roles in determining A. sexdens demography.