Assuming that the period 2.75-2.65 Ga corresponds to a single, but global, geodynamic event, we investigate-through numerical experiments-the mechanisms that could have led to the profound continental reworking that occurred at that time. Although the extent of the crisis at the Earth's surface pledges in favour of the involvement of mantle plumes, our numerical experiments suggest that the thermal impact of mantle plumes is unlikely to explain both the amplitude and timing of the thermal anomaly, as observed in the Superior Province (Canada) and the Yilgarn Craton (Australia). Similarly, moderate crustal thickening can not lead to significant reworking of the continental crust within the observed time constraint. Crustal thickening with a factor greater than or equal to 1.5 is also unlikely because it is not consistent with the moderate metamorphic grade observed at the surface of many Archaean cratons. Burial of a radiogenic crust under a 10 km thick greenstone cover also falls short of explaining, not so much the amplitude and the extent, but the timing of the thermal anomaly. In contrast, the combination of the thermal anomaly related to the greenstone blanketing effect with the heat transfer from a plume head spreading at the top of the thermal boundary layer can adequately explain the amplitude, the timing, and the extent of the 2.75-2.65 Ga crisis.Our favoured model involves a global rearrangement of convection cells in the deep mantle and formation of multiple mantle plumes. The greenstones emplaced at the surface and the plumes that spread in the thermal boundary layer contributed to heat the crust from both above and below. This produced massive crustal partial melting that reached its climax ca. 40 Myr after the emplacement of the plumes and associated greenstone cover rocks. This led to gravitational instabilities in the crust, as dense greenstone cover rocks began to sink into the thermally softened crust and granite domes rose in response. The extraction of heat-producing elements toward the upper part of the crust has contributed to the cooling and stabilisation of the cratons. This succession of events, which is not incompatible with plate-tectonic processes, may have profoundly changed the nature of the crust exposed at the surface and could explain the contrasting geochemical signatures of Archaean and post-Archaean shales. (C) 2003 Elsevier B.V. All rights reserved.