The 2008 Mw = 6.4 Movri earthquake ruptured a NNE right lateral strike-slip fault about 30 km south of the city of Patras. Although some strike-slip activity on minor faults was known, there was no tectonic evidence of large scale NS striking fault and such a large event was not anticipated. Following the event, a network of six stations was installed for 4 months in the epicentral area in order to monitor aftershocks and in particular the northern part of the rupture area closest to the city of Patras. We combine these new aftershock observations with GPS measurements of an already existing geodetic network in the area performed just after the earthquake, as well as with SAR interferograms, together with already published source studies, in order to refine already proposed models of this event. The combined data set allows defining much more accurately the lateral and vertical limits of the rupture. Its length inferred from geodesy is ∼15 km and its modelled upper edge ∼17 km. The seismic moment then constrains the lower edge to coincide, within a few kilometres, with the Moho interface. The absence of seismicity in the shallow crust above the co-seismic fault is interpreted as a result of the decoupling effect of possible presence of salt layers above the rupture area, near 14 to 16 km in depth, which favours our interpretation of an immature strike-slip fault system, compatible with the absence of surface ruptures. The immature character of this large crustal fault is further suggested by the high variability of focal mechanisms and of fault geometries deduced from aftershock clusters, in the strike direction. Its geometry and mechanism is consistent with the crustal shear, striking NNE, revealed by GPS in this region. This shear and faulting activity might be generated by the differential slip rate on the subduction interface, 50 km to the south, leading to a north-northeastward propagating strike-slip fault zone. The wide extension of the aftershock distribution forming a NNE alignment, beyond the rupture area towards the north, suggests a localization process of the shear strain, which could be the preliminary stage of fault propagation further to the NNE. An alternative speculative model for this regional stress could be the existence of a well-developed NNE striking shear zone within the uppermost mantle, marking at depth the southward propagation of the northern branch of the North Anatolian fault. Both models may not be exclusive of each other, and in fact their sources may be mechanically interdependent.