The peak of high-temperature, low-pressure anti-clockwise granulite facies metamorphism in the northernmost part of the Northern Marginal Zone of the Limpopo Belt has been dated at 2591¿4 Ma using conventional U-Pb zircon geochronology on a syntectonic charnockite intrusion. The thermal peak of 800 ¿C at 0.5 GPa reflects a strong perturbation of the geotherm and was reached during compressional deformation and coincides with vast granite intrusion into the overlying craton. The coincidence of the temperature (T) maximum and compression is thought to be a response to lithospheric mantle thinning that resulted in granite emplacement and a consequent change in upper crustal rheology. After initial exhumation and cooling, these high-temperature, low-pressure rocks resided at a comparatively shallow depth (≈10 km) for 620 Ma. Field, microstructural, and geochronological investigation of a low-temperature overprint establishes that reworking and final exhumation occurred at 1.97 Ga along discrete greenschist facies thrusts. The 40Ar/39Ar laser ablation of hornblende and biotite and Pb-Pb step leaching of sphene are shown to be useful geochronometers in dating the low-grade assemblages. Our results indicate that early Proterozoic deformation lasted 10 to 20 Ma. However, a number of ages intermediate between the events at 2.6 and 2.0 Ga have been reported for the Northern Marginal Zone. These include Pb-Pb garnet, Ar-Ar step heating amphibole, and U-Pb zircon ages. Such ages suggest that these minerals were partially rejuvenated below their commonly believed closure temperature. Reinterpreting these intermediate ages as partial overprint or inheritance was possible in the Northern Marginal Zone due both to the large time gap between the two tectonometamorphic events and the use of a multimethod multichronometer approach. These observations call for great caution when extrapolating mineral ages into cooling histories, especially so when the resulting cooling is slower than thermally and/or mechanically possible and suggests that the geochronometry in other regions that underwent multiple tectono-metamorphic events may need to be reassessed. ¿ American Geophysical Union 1996