AbstractIn the Miocene (23–5 Ma), a large wetland known as the Pebas System characterized western Amazonia. During the Middle Miocene Climatic Optimum (c. 17–15 Ma), this system reached its maximum extent and was episodically connected to the Caribbean Sea, while receiving sediment input from the Andes in the west, and the craton (continental core) in the east. Towards the late Miocene (c. 10 Ma) the wetland transitioned into a fluvial-dominated system. In biogeographic models, the Pebas System is often considered in two contexts: one describing the system as a cradle of speciation for aquatic or semi-aquatic taxa such as reptiles, molluscs and ostracods, and the other describing the system as a barrier for dispersal and gene flow for amphibians and terrestrial taxa such as plants, insects and mammals. Here we highlight a third scenario in which the Pebas System is a permeable biogeographical system. This model is inspired by the geological record of the mid-Miocene wetland, which indicates that sediment deposition was cyclic and controlled by orbital forcing and sea-level change, with environmental conditions repeatedly altered. This dynamic landscape favoured biotic exchange at the interface of (1) aquatic and terrestrial, (2) brackish and freshwater and (3) eutrophic to oligotrophic conditions. In addition, the intermittent connections between western Amazonia and the Caribbean Sea, the Andes and eastern Amazonia favoured two-way migrations. Therefore, biotic exchange and adaptation was probably the norm, not the exception, in the Pebas System. The myriad of environmental conditions contributed to the Miocene Amazonian wetland system being one of the most species-rich systems in geological history.