Sedimentologic data and palynological ages from Paleogene clastic deposits of the northern and central Altiplano plateau suggest foreland basin development in the central Andes by mid-Paleocene time. The nonmarine Potoco Formation (3000-6500 m thick) constitutes the majority of Cenozoic basin fill. The Potoco overlies the Santa Lucia Formation (50-300 m thick), previously dated as mid-Paleocene by mammal fossils and magnetostratigraphy. New geochronologic data for the Potoco Formation include late Eocene to Oligocene palynomorph assemblages recovered throughout lower to upper stratigraphic levels. These ages and published 40Ar/39Ar and K-Ar ages from overlying upper Oligocene-lower Miocene volcaniclastic rocks indicate (1) nondeposition or greatly reduced deposition (average sediment-accumulation rates <10 m/m.y.) from mid-Paleocene to middle Eocene time (top Santa Lucia to lowermost Potoco Formation), followed by (2) rapid deposition (average sediment-accumulation rates up to 500 m/m.y.) throughout late Eocene and Oligocene time (majority of Potoco Formation). Lithofacies and paleocurrent data confirm both depositional phases. A 20-100-m-thick interval of superimposed paleosols in the basal Potoco supports reduced sediment accumulation and low rates of subsidence during mid-Paleocene to middle Eocene time. The overlying main body of the Potoco contains facies assemblages and stratigraphic architecture (dominantly nonerosive fluvial sheet sandstone interbedded with overbank mudstone) consistent with an aggradational fluvial system in a rapidly subsiding late Eocene-Oligocene environment. An upsection reversal in paleocurrents across the Santa Lucia-Potoco contact, from west-directed to east-directed patterns, also indicates a major change in depositional systems. The switch from low to high rates of subsidence and reversal in paleocurrents can be explained by eastward migration of a foreland basin system. Basin migration may have induced a change from a mid-Paleocene-middle Eocene, broad-wavelength, low-amplitude forebulge to a late Eocene-Oligocene, rapidly subsiding foredeep. Advance of such a forebulge-foredeep pair would require a topographic load to the west that propagated eastward through time. Although synchronous loading east of the Altiplano is plausible, paleocurrent data suggest a greater influence of sediment dispersal systems derived from a belt to the west. In general, the Paleogene sedimentary history of the Altiplano is compatible with shortening and crustal thickening in an eastward-propagating contractional belt to the west in the modern arc and forearc regions of westernmost Bolivia and Chile.