Geochronological data for timing of Gulf of Corinth rift-margin fault migration is presented from the Megara basin, precursor to the active eastern Corinth rift. Fifteen sanidine separates from a thin tuff near the top of the sedimentary basin fill analysed by laser-fusion give apparent 40Ar–39Ar ages from 2.78 to 4.53 Ma. A weighted mean age of 2.82 ± 0.06 Ma (at 2s error) for the three youngest aliquots (range of mean ages 2.78–2.86 Ma) is regarded as a maximum probable age for ash eruption. Together with magnetostratigraphy results, this age constrains timing of Megara basin abandonment and likely propagation of the active South Alkyonides coastal faults that presently bound the southern rift-margin to the late-Pliocene, 2.2 Ma. Initiation of the coastal faults caused uplift, fluvial incision and calcrete formation on geomorphic surfaces over the abandoned Megara basin fill. A petrocalcic laminar horizon from a supermature calcrete unconformably capping the fill gives a U–Pb age of 0.77 ± 0.08 Ma, dating a late stage in the history of calcrete development. The new age for initiation of active faulting in the eastern rift yields a low estimate of long-term mean extension along the active South Alkyonides coastal faults of 0.9–1.4 mm a- 1, consistent with previous geological data. This rate is less than the 100-year GPS-determined geodetic extension rate of 6 ± 2.7 mm a- 1 measured along a 23°E meridional array just west of the Alkyonides gulf. It implies either that the geodetic rate declines rapidly over the 15 km distance into the eastern gulf or that the geodetic rate is unchanged but extra strain is taken up aseismically and/or along antithetic and unrecognised major intrabasinal faults. In the full graben of the central rift displacements are 3.5 km along the Xylocastro and Antikyra faults; the mean long term extension rate here since 2.2 Ma is 3.5 mm a- 1, much less than current geodetic rates of 10 mm a- 1. In the western rift, despite a lack of precise chronological data, geodetic and extension rates seem comparable. Overall, a late-Pliocene to early Pleistocene age is likely for initiation of the deep-marine Corinth rift, with no evidence for rift propagation, either eastwards or westwards. More generally, our results constrain timing of strain localisation and vertical axis rigid block rotation over the Aegean–Anatolian plate and demonstrate that intraplate deformation can be accomplished rapidly in response to regional-scale tectonic drivers.