Abstract The Eocene–Miocene Mesohellenic Trough is an elongate sediment-filled tectonic basin trending NW across central Greece and into Albania. Neotethyan oceanic rocks, including Triassic–Jurassic rift-related volcanic rocks and deep-sea sediments, accretionary mélange and ophiolitic complexes, crop out along its margins. These units were tectonically emplaced onto the Pelagonian microcontinent to the east and the Apulian–African continental margin to the west. In northern Greece, the mid-Jurassic Vourinos ophiolite on the eastern margin of the trough is geographically separated from the synchronous Pindos ophiolite along the western margin by a minimum c. 20 km distance. The sedimentary fill of the trough obscures their presumed subsurface continuation, although magnetic surveys identify thick ‘ophiolitic’ rocks beneath the basin. We interpret these ophiolites as parts of the same oceanic slab, two parts of a single larger oceanic complex we now term the Mesohellenic ophiolite. Comparable ophiolitic complexes to the south (the Koziakas and Othris) and the ophiolites of the Mirdita complex to the north in Albania are considered as members of this same complex. Geological and petrological data from the Vourinos and Pindos ophiolites define intra-slab heterogeneity. Vourinos essentially is a ‘Penrose-style’ ophiolite with ‘supra-subduction’ compositions; the less continuous Pindos ophiolite shows coexisting mid-ocean ridge basalt and island arc characteristics. Ophiolitic rocks that seem to represent geographical overlap between these characteristic associations crop out along their northern (Dotsikos strip) and southern (Mesovouni) margins. Variations in mantle strain conditions across the ophiolitic slab have been mapped, and demonstrate a single orientation of deformation; this is explained by variable strain kinematics that persisted across the ductile–brittle boundary. A continuity from ductile to brittle emplacement structures spans the Mesohellenic Trough, independent of petrological association, and indicates the original relative positions of these ophiolites within the oceanic slab. These structures illustrate tectonic ‘steps’ of obduction from the ridge crest onto the Pelagonian margin to the east, and can be relatively timed by the overlap of magmatism with ductile deformation in different parts of the slab. Hence, rotations of original horizontality are dated to the period preceding cessation of ductile field deformation, while still in the oceanic environment. The morphology defined by these structures and the horizontal rotation of stratigraphy are analogous to a spoon- or scoop-shaped nappe originating in the ductile field at its base, and crossing into the brittle field rapidly at its leading edge (Vourinos), whereas the mylonitic deformation characterizes the ‘trailing’ end (Pindos). Age relations require that geochemical variation between the two complexes must be explained within a model of synchronous generation, possibly with apparent ‘supra-subduction zone’ rifting of originally heterogeneous mantle, or an overlapping series of diverse processes of magma generation in an initially homogeneous mantle. Indications of the original ridge crest directions suggest the operation of several simultaneous spreading centres, separated by transform faults or ‘pseudofaults’. A palinspastic reconstruction of the slab constrains applicable oceanic models and provides the basis of future research.