The Uatumã event is a silicic large igneous province (SLIP) covering an area of 1.500.000 km2 of the Amazonia craton at ca. 1890–1850 Ma. New paleomagnetic data and U-Pb apatite ages for coeval microgranitic, andesitic, and basaltic dykes in the Carajás Province (southeastern Amazonia craton) are presented. Two primary characteristic remanent magnetizations (ChRMs) were isolated for the Carajás dykes which were combined with published results for coeval rocks in the São Felix do Xingu area, producing two grand mean poles: (1) “CA1”, dated at 1880 Ma on zircon and apatite, whose mean site directions cluster around the mean Dm = 132.8°, Im = 76° (N = 26, α95 = 5°, k = 32.7), yielding the paleomagnetic pole at 328.7°E, 23.3°S, (A95 = 8.7° K = 11.7), and (2) “CA2” from well-dated microgranitic dykes of the Velho Guilherme Suite (1855 Ma U-Pb on zircon and apatite). Site mean directions group around the mean Dm = 240.3°, Im = 11.8° (N = 18, α95 = 10.8°, k = 11.1), which yielded the paleomagnetic pole at 221.3°E, 30.2°S, (A95 = 8.8°, K = 16.2). The remanent magnetization of the Velho Guilherme microgranitic dykes is carried by pseudo-single domain (PSD) magnetite. Its primary origin is supported by a positive baked contact test. Two younger secondary components were also isolated in the Carajás dyke swarm: one probably associated to the Mesozoic Central Atlantic magmatic province (CD3), and another component (CD4) with yet undefined age. A new configuration for the Amazonia and West Africa during the Paleoproterozoic, which aligns the Sassandra shear zone (in Africa) and the North Guiana Through and other shear zones (in Guiana Shield), is supported by geological and paleomagnetic data. The large angular distance between the 1880 Ma CA1 and 1855 Ma CA2 (Q ~ 6) poles that are relatively close in age implies in a high plate velocity for the Amazonia craton, which is not consistent with modern plate tectonics. A similar large discrepancy between 1880 and 1860 Ma poles was also observed at least as many as six cratons. Although other hypotheses can be invoked, such as non-uniformitarian magnetic fields, the occurrence of large-scale true polar wander at this time may be the most parsimonious explanation of the widespread discrepancies.