The origin of the transcontinental Amazon drainage system remains unrevealed. Sedimentary deposits formed from the Neogene in the Amazonas and Solimões Basins constitute natural archives for reconstructing this event in space and time. However, paleoenvironmental and chronological analyses focusing on these deposits, or even their basic mapping, are still scarce to allow such investigation. In this context, primary interests are fluvial strata related to the lithostratigraphic Içá Formation, mapped over a widespread area in western Amazonian lowlands. Although long regarded as Plio-Pleistocene in age, this unit has not yet been dated and its overall depositional setting remains largely undescribed. The main goal of the present work is to contribute for improving facies analysis and chronology of these deposits, approaching an area in southwestern Amazonia and another in northern Amazonia, which are located more than 1000km apart. Despite this great distance, the sedimentological and chronological characteristics of deposits from these two areas are analogous. Hence, facies analysis revealed paleoenvironments including active channel, abandoned channel, point bar, crevasse splay and floodplain, which are altogether compatible with meandering fluvial systems. Similarly, optically stimulated luminescence (OSL) dating revealed thirty three ages ranging from 65.4±16.9 to 219.6±25.1ky (in addition to three outliners of 54.0±7.6, 337.3±36.9 and 346.6±48.6ky), and nine 97.1±9.9 to 254.8±23.8ky for the areas in southwestern and northern Amazonia, respectively. These data lead to establish that deposits mapped as Içá Formation over a vast area of western Brazilian Amazonia have a Mid-Late Pleistocene age, rather than the previously inferred Plio-Pleistocene age. It follows that if Plio-Pleistocene deposits exist in this region they remain to be dated and must be restricted to a narrow belt in western Amazonia, as well as isolated occurrences underlying the Mid-Late Pleistocene strata characterized herein. The combination of data from this work with previously published provenance studies supports main Andean sediment sources only in the Mid-Late Pleistocene. It is proposed that before this time, the Amazon River was restricted to eastern Amazonia, being separated from western Amazonian drainage basins due to the presence of the Purus Arch. Erosion and/or subsidence of this geological feature would have promoted the connection of these drainage systems, ultimately with the expressive record of the transcontinental Amazon pathway into the Atlantic Ocean in the Mid-Late Pleistocene.