{During the summer of 1993, a network of seismological stations was installed over a period of 7 weeks around the eastern Gulf of Corinth where a sequence of strong earthquakes occurred during 1981. Seismicity lies between the Alepohori fault dipping north and the Kaparelli fault dipping south and is related to both of these antithetic faults. Focal mechanisms show normal faulting with the active fault plane dipping at about 45° for both faults. The aftershocks of the 1981 earthquake sequence recorded by King et al. (1985) were processed again and show similar results. In contrast, the observations collected near the western end of the Gulf of Corinth during an experiment conducted in 1991 (Rigo et al.), and during the aftershock studies of the 1992 Galaxidi and the 1995 Aigion earthquakes (Hatzfeld et al. 1996; Bernard et al. 1997) show seismicity dipping at a very low angle (about 15°) northwards and normal faulting mechanisms with the active fault plane dipping northwards at about 30°. We suggest that the 8–12 km deep seismicity in the west is probably related to the seismic—aseismic transition and not to a possible almost horizontal active fault dipping north as previously proposed. The difference in the seismicity and focal mechanisms between east and west of the Gulf could be related to the difference in the recent extension rate between the western Gulf of Corinth and the eastern Gulf of Corinth, which rotated the faults dipping originally at 45° (as in the east of the Gulf) to 30° (as in the west of the Gulf).}

We report here the results of a tomographic lithospheric study in the area of the Corinth and Evvia rifts (Greece), designed to constrain the mechanism of continental extension. Sixty seismological stations were deployed in the area for a period of 6 months, and 177 teleseismic events were recorded by more than five stations and gave more than 2000 travel time residuals (P and PKP phases), which were inverted to image the velocity structure down to 200 km depth. We use both a linear and a nonlinear method to invert the data set. The main result is a long-wavelength positive velocity anomaly located in the upper mantle, which is interpreted as the subducted African lithosphere. The subducted lithosphere is well defined from ∼7O km depth down to 200 km. From synthetic tests as well as from the amplitude of the anomaly (more than +7%) we conclude that the subduction continues below 200 km. In addition, a second positive velocity anomaly of about +4% from the surface down to 40 km depth, located north of the Gulf of Corinth, has been found. This is interpreted as the result of a crustal thinning of several kilometers (∼5 km), shifted to the north from the Gulf of Corinth and trending obliquily NW-SE. We suggest that this crustal thinning is mainly related to the Miocene widespread extension in the Aegean and that the Quaternary Corinth rift initiated where the crust was already thinned. The different styles of deformation of the eastern and western part of the rift are consistent with this interpretation. No clear velocity anomaly can be related to the Evvia rift.